Self Proclaimed Die Expert

Virtual Stoning

noreply@blogger.com (eric kam) — Tue, 13 Aug 2019 13:03:00 +0000

Not sure if I should feel guilty "Taking the piss" online in the comments section of a sheet metal forming group post on LinkedIn, wherein my replacement in managing the blog at AutoForm makes some rather challenge-able statements on the value of Virtual Stoning in sheet metal FEA results.

only two misleading or vague explanations in the response..

It might be petty, but the response does little to explain what Virtual Stoning is and even opens up the chance for severe misunderstanding about what stoning in real life and and what Virtual Stoning is. OK, it is REALLY REALLY petty, but I can't help myself.

Stoned panel, areas B, C, D the stone leaves marks in areas where the metal lifts the flat stone higher than adjacent areas

When applied in practice, stoning is the process where die tryout personnel use abrasive stones to lightly abrade the surface of a sheet metal panel after it has been removed from the stamping dies to ascertain if there are any areas where the curvature of the deformed part deviates from the intended curvature. The stones are normally relatively small like the size of a pocket whetstone and made of similar material. When rubbed over the part with the stone aligned perpendicular to the curve of the sheet metal then slid up and down following the curve it would inevitably scratch the surface of the panel. In areas where the sheet metal was dimpled into or out from the surface of the sheet there were telltale signs that the stone was rubbing off more of the coating in some areas compared to others.

Stoning might detect two types of defect in the formed part:

dimples or pimples caused by trapped debris in the tool between the sheet and tool surface
localized curvature changes resulting from springback recovery that makes the metal want to expand from the shape held in the die then bowing up or down from that intended shape

Localized pockmarks caused by retained slivers of trimming burrs or other debris in the die could be isolated in this manner and the tool makers could recondition the die surface with scotchbright pads or other mild abrasives to remove the offending articles from the dies punch surface.

In other uses of the technique localized springback distortions like localized "highs" or "lows" might also cause more widely spread areas where the stone might bridge over areas that are due to compressive springback behavior bulging up or down from intended curvature. When this was detected in the forming of the parts within a die (or as a result of a secondary forming process like hemming or welding) there was little that could be done in production tools to resolve it. The metal forming process itself might have to be adjusted to mitigate or eliminate the source deformation issue that resulted in the distortion.

It should be noted that no metal forming simulation today can predict or resolve the first mode. As there is no dirt or slivers resulting from trimming in simulation. The second mode could be resolved if the part might be formed in a way to reduce the negative compressive stresses during forming that make the sheet expand in springback.

I do hope that they get it right and are able to better understand that which their blog intends to inform people on. A little Socratic passive aggressiveness should only help their (my former) blog get better.

Disclosure: PAM-Stamp in addition to LS-Dyna and the comments Author's software all have similar function to emulate the effect of stoning. Most are not doing so in Virtual Reality when it comes to Stoning. However, the severity of such defects on the aesthetics of the parts are viewable in VR using PAM-Stamp

Class A Surface panels and predicting "highs/lows"

noreply@blogger.com (eric kam) — Mon, 25 Mar 2019 19:57:00 +0000

Reflected lights in a simulated sheet metal stamping

The detection/prediction of Class A panel surface quality was, during my sheet metal stamping heyday, something of a specialty; at least the debunking of and explaining of the phenomenon and mitigation/correction of it. Quite often needing to tell others that you can't simply compensate your way out of it. https://kam-stampingguru.blogspot.com/2009/10/springback-compensation-part-2-of-many.html

Numerical simulation color plot indicating distortion normal to sheet, indicator of what?

What always made the prediction of the class A panels springback distortion turn into an "It depends" discussion was the fact that for contour plots like the one above, is the lack of any standard to indicate that a given value threshold was considered "good" or "bad". Unlike the prediction of splits or wrinkles there is no simply quantified way to determine for a unique sheet metal shape if the distortion of the shape would render the part ugly or not.

"Stoned" panel in sheet metal indicating what also?

The method of stoning was something that we would do in the plant to make very rapid and moderately objective observations regarding the surface quality that the die, process, and product all together generates. The method required us to take an abrasive stone of predetermined size usually about 4 inches long and rub the flat face of the stone against the sheet from front to back (in the case of this door) or left to right (for horizontally mounted panels like hoods and roofs). Areas with severe changes in curvature would invariably get highlighted as the stone would bridge the differing curvatures and rough the adjacent sheet metal more or less. But this did not always indicate the nature of the issue. The Stoning image in grey above correlates well to the color plot above, but the significance of this is not clear, since if I use different sized stones or rub at different angle the results would appear to change.

An inspector evaluates the contour of the panel under fluorescent lights to see if it is "correct"

The other method would be to shine the panel with oil and place in front of or underneath an array of fluorescent lights. The trick here was to observe if the lights were being reflected in the oil in a consistent manner. Did the straight light tubes reflections remain straight? Do they appear to diverge or converge in ways that seem to defy the planned curve of the panel?

This method is very hard to capture in numerical stamping simulations, since interpretation of color scale plots do not always represent the visual aesthetic (rarely to be precise). Additionally, most indirect interpretations leave open the potential to over-engineer some parts and under-engineer others. Recalling that there is NO WAY TO ELIMINATE springback, we might find after multiple engineering reviews that we have obtained the best result possible when the defect is as minor as possible, but it will still be there.

From my presentation at the NVIDIA GTC

If we could repeatably review the planned sheet metal parts based on numerical simulations using the intended method of surface quality approval, then we could limit our exposure to risk by knowing what the best possible result is and if our reality matches close enough that result. Observing the springback effects in VR is a novel, albeit maybe not entirely practical, approach to evaluate Class A panels. Much better than trying to guess based on numerical parameters that we correlate with good/bad surface. Even better than predicting it, in VR we can really stand and review the panels in true scale and from real vantage points to decide not only if a defect is likely but more importantly if it matters.

Reflections on Formability Analysis

noreply@blogger.com (eric kam) — Mon, 07 Jan 2019 17:21:00 +0000

EGO! Oh the ego!

More than two years after abandoning sheet metal stamping simulation as a trade and moving onto engineering for "Human-Centric Product and Production Planning", I can say that our egos when we were promoting solutions for sheet metal stamping simulation knew no bounds. I still see it today when discussing with folks "Industry 4.0" and "Smart Factories" how we pretty much convinced ourselves that everything revolved around the stamping feasibility of sheet metal parts.

That is a survival tactic, for sure, but we pretty much held a straight face when promising people that if they ran simulations with [insert code here] that we could assure them smooth production with minimal surprises. Anytime that the results in reality did not match the predictions, we could always safely land with the statement: "Well the [insert property or input boundary condition here] that you used in the simulation is not the same as the delivered [insert property or input boundary condition here] used in production.

That's not even evasive or misleading, since at the time of simulation nearly every aspect of the simulation will have to be differently characterized than how things work in real world. But execution of the actual stamping goes far beyond just the color scale plots of the stamping, or casting, or weldment.

Where in my previous impression of the smart factory did we consider the effect of the non-tooling production planning. How can we resolve the human-centric issues in building and assembling the products of the future.

Virtual Reality: My new calling....

noreply@blogger.com (eric kam) — Tue, 30 Aug 2016 21:38:00 +0000

I have begun my transition into a new career outside the realm of sheet metal stamping. I now work at a Virtual Reality Engineering company and am shifting my attentions toward supporting a broader slice of the new product design and engineering value stream. And I started a NEW BLOG.

ESI-Group's IC.IDO is the first industrial grade Virtual Reality solution.

The number of applicable use-cases for virtual reality during engineering and design of new products is seemingly boundless. Already our customers have seen great success when implemented as part of their design, engineering, and packaging processes.

But there have been also early adopters who are using IC.IDO Virtual Reality to improve their sales processes and customer support.

They are using VR for training their customers and filed service technicians, when deploying new products.

I will continue blogging on technology. But less and less on this site and more on my new BLOG at http://kam-ono-sendai-7.blogspot.com/

Goodbye, Self-Proclaimed-Die-Expert

noreply@blogger.com (eric kam) — Fri, 29 Jul 2016 18:40:00 +0000

After 22 years in the sheet metal stamping and metal formability field, this self-proclaimed-die-expert (actually THE Self-Proclaimed-Die-Expert) is hanging up his stamped sheet metal spurs.

“I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve.”- Bilbo Baggins (J.R.R. Tolkien, The Fellowship of the Ring)

I has been a pleasure serving the sheet metal stamping industry for all these years:

2 years as a technical trainer with General Physics
2 years as Senior Engineer with Forming Technologies Inc.
2 years as a Senior Engineering Consultant with Performance Innovations, Inc.
6 years as Principal Engineer of KamConsulting
14+ years with AutoForm Engineering (first 4 years as KamConsulting, 2 years as Technical Services Manager, 8 years as Product Manager of Solution for Design, Bidding&Planning Solution, HemPlanner*, Incremental*, ThermoSolver*, Sigma*, Springback Compensator* (* denotes products that I had been just an interim product manager for, now REAL product managers perform those roles)

I will leave this here, but will turn off comments and will not add or update any information to this blog. Sheet metal is dead to me.

I am embarking on a new adventure in a field yet to be announced.

I will start a new blog where my new adventures will be recorded for all 3 of you who care to read it.

http://kam-ono-sendai-7.blogspot.com/2016/07/ready-player-one.html

May 2016 Metalforming Simulation and Stamping & Dies Technical groups: Tribology Simulation for stamping

noreply@blogger.com (eric kam) — Thu, 12 May 2016 09:03:00 +0000

The May 2016 meeting of the Stamping and Dies Technical Group and the Metalforming Simulation Technical group was a very interesting discussion lead by Dr. Johan Hol, PhD. The topic of friction during metal forming simulation was the topic.

A single area in the die will display different friction performance

I have posted before that friction is one of the biggest "black boxes" when considering metal forming simulation. FEA codes attempt to describe the complex system of interactions between sheet metal, lubrication, and tool with a single number. This single number is often assumed to be constant, isotropic, and homogeneous. But is that a valid assumption? The question has been posed several times here and in other blogs.

Triboform did an admirable job sharing how they address these challenges (and more than we had even considered in previous posts) during their presentation during the May meeting. The recording is posted below.

https://youtu.be/B-2ppYlcPEs

Lubrication for simulation challenges

noreply@blogger.com (eric kam) — Thu, 17 Mar 2016 14:19:00 +0000

source: thefabricator.com

Among the common questions asked by users of sheet metal forming simulation is "What friction (lube) value should I use?".

What follows that question is a little bit of a tap dance, improvisational theater, creative thinking, and other non-answer answers. I've had to answer that question with things like: If the die is polished Cast Iron, and the blank is EG, and using mill oil only; 0.15 is reasonable. If you chrome the tools maybe 0.12. You wrap the parts in garbage bags and drown them in whale snot; 0.08. However, these suggestions are tricky as the measurement and performance of the lubes in question under varying circumstances can be nearly impossible to pin down.

Simple coulomb friction model

In order to run a stamping simulation you will need to enter a friction number. That number is supposed to represent the effective friction between the tool and sheet. It should encompass how the tool surface, its coating, its lubrication interacts with the sheets surface, coating, and lube. We enter a single number.

That number is a simple coulomb friction coefficient. A single number can only represent a single condition, but even in high school physics we learn that we need to consider more situations like static friction versus kinetic friction; the friction needed to overcome to start an object sliding versus the friction once the object is already sliding. However, the metal forming tribological system is even more complex that a single coulomb friction would imply. The lubrication used will create different coefficients depending on the forming pressure application and/or the velocity at which the material is sliding in the die.

Friction in a die has potential different coefficients of friction in each area, as pressure and velocity affect lube performance

Describing the lubrication situation with a single number is an efficient solution, but one that has the potential to over/under prescribe the effect of the lubrication. The lubrication between the sheet metal and the tool, will change in its behavior-hydrostatic, hydrodynamic, boundary, or elasto-hydrodynamic. It is possible that under higher pressure the lubrication will work more/less effectively and decrease/increase the effective coefficient. Likewise, if the velocity increases or decreases we might note an increase or decrease in coefficient.

It is possible in finite element analysis to define more complex relationships between the applied pressure and observed velocity in order to use the more precise definition of friction. In AutoForm software the user can define pressure dependency and velocity dependency. Simply by defining the reference pressure and reference velocity along with the pressure exponent and velocity factor. This enable the application to for each time step use an appropriate instantaneous friction value to determine the deformation effects on sheet metal. This assumes however that we have a reasonable value for friction coefficient that has been somehow measured and how pressure and velocity effect that friction.

With appropriate inputs, pressure and velocity dependence can be defined, but where does that information come from?

But quantifying a friction coefficient is not so simple. Different friction tests produce highly different friction coefficients for the same materials. There are many challenges to measuring the friction coefficients in these complex systems. Research on this topic still leaves the topic open for interpretation.

Recent research and release of new software based on that research show interesting potential. Triboform B.V. is a new company with new solution to defining the behavior of friction values for sheet metal stamping simulation. With this new technology, users can define the material of the die and sheet, along with a lubrication and simulate the friction for a range of pressures and velocities. This way arriving at a range of friction coefficients which can be used by commercial stamping simulation codes to define the friction for every element and every time step to recognize the effects of pressure and velocity on the friction effects.

source: http://www.triboform.com/

This May the Stamping and Dies Technical Group of SME will host a webinar featuring guests from Triboform, who will share with us a short technical discussion on their technology and solution.

Stamping and Dies Tech group and Metal Forming Simulation Tech Group meeting

Wednesday May, 11, 2016
1:00 PM Eastern Daylight Savings Time (US)

Meeting number: 649 116 325
Meeting password: Stamping
https://smeweb.webex.com/smeweb/j.php?MTID=m2bb6c12127c229bb4dab2133493b23f9

Access code: 649 116 325

Audio connection:

1-877-668-4493 Call-in toll-free number (US/Canada)
1-650-479-3208 Call-in toll number (US/Canada)
Global Call-in numbers

The "Stamping System"

noreply@blogger.com (eric kam) — Fri, 12 Feb 2016 20:37:00 +0000

Many years ago when we were planning one of several courses developed on the topic of metal forming analysis and sheet metal stamping process control, my colleagues and I began to think about the methods of control of a system. In order to define and control and system it was essential to be able to define the inputs and the outputs of a system.

One useful analogy was to compare metal forming process control to that of dieting and trying to control ones weight. We know the inputs: Food, exercise and activity, genetics. We know the outputs: our weight. If we monitor our inputs, and monitor our outputs, we collect a lot of data. but that data does not correlate until we can assign a relationship between the inputs and the outputs. This we could think of as a transformation characteristic.

The control mechanism, or transformation characteristic, is calories. How many calories of fuel do we take in, how many calories of energy do we burn? That information along with the output of the scale we might use to monitor our weight is when we can start to really try to make some gains in the control of the system. If we are not using a measurement like calories, we won't get a handle on our weight gain or loss.

Weight watchers is just another spin on that idea where the use of points replaces calories, but still to same effect.

In sheet metal stamping we use STRAIN as the transformation characteristic. That links the inputs to the outputs, and through the measurement and prediction of strain in sheet metal stamping we aim to better understand how our products and processes will perform.

In a recent blog post by AutoForm Engineering, they introduced a short list of the input parameters of a simplified stamping system. A Topic we have delved into here from time to time.

What can really be expanded upon is to take the short list of input parameters from the AutoForm Blog, and really expand upon them to better understand the relationships of them to the outputs we observe. And to better plan for which ones we control, which ones we "think" we control, and which ones are outside our control.

Blank/Coil	Production Line	Die
Thickness	Applied ram tonnage	Tooling radii (design intent vs. maintenance)
Yield and tensile strengths	Ram speed	Die surface and coating
Coil width and pitch	Applied cushion tonnage	Drawbead geometry (if any)
n-value and R-value	Shut height settings	Pad/binder spotting
Coating (if any)	Lubrication (if any)	Tool and surface condition
Mill oil	Cushion pin placement	In-die pressure systems
Edge condition of blanks	Parallelism and maintenance	Equalizer blocks or lack of
Weld line/patch weld/tailored thickness	Drive type	Tooling parallelism
Grain and rolling direction	Automation feed rate	Temperature
Substrate metallurgy	Blank/coil alignment to die	Location on bolster/ram
… and probably more	… and probably more	… and probably more

Input parameters re-posted with permission by AutoForm Engineering

2016 SME Stamping and Dies Technical meeting schedule

noreply@blogger.com (eric kam) — Mon, 11 Jan 2016 17:16:00 +0000

We have scheduled the 2016 monthly web meetings for the Stamping and Dies Technical Group under the Forming and Fabricating Technical Community. The meetings will be as they have been for the last several years.

The Second Wednesday of each month. Beginning Wednesday January 13, 2016. 1:00 PM EST

Join WebEx meeting
Meeting number: 649 116 325
Meeting password: stamping

Join by phone
1-877-668-4493 Call-in toll-free number (US/Canada)
1-650-479-3208 Call-in toll number (US/Canada)
Access code: 649 116 325
Global call-in numbers | Toll-free calling restrictions

We follow a loose calendar for topics each month and welcome all SME members as well as members of partner associations such as PMA and FMA to join in the discussion each month. You are welcome to contact us to suggest you ideas for future topics and we welcome suppliers, vendors, and technology companies to share their advances with us.

Month	Day	General Topic (Theme)
Jan	13	Stamping Fundamentals
Feb	10	Materials Topics (tool, sheet, etc)
Mar	09	Press room topics
April	13	Topic within F&F (broaden scope)
May	11	Design and engineering
June	08	Education and Development - mid year review
July	13	Stamping Fundamentals
Aug	10	Materials Topics (tool, sheet, etc)
Sept	14	Pressroom
Oct	12	Topic within F&F
Nov	09	Metal Forming Simulation and FABTECH
Dec	14	Education and development- Year in Review?

Can't join the meeting? Contact support.

IMPORTANT NOTICE: Please note that this WebEx service allows audio and other information sent during the session to be recorded, which may be discoverable in a legal matter. You should inform all meeting attendees prior to recording if you intend to record the meeting.

This engineers opinion on stress, it is not what is "put upon you"

noreply@blogger.com (eric kam) — Thu, 17 Dec 2015 12:23:00 +0000

In this TED talk psychologist Kelly McGonigal she outlines a research study that suggests that stress may only be bad for you if you believe that to be the case. That it is how we react to stress that is the issue.

Source: TED Talks

But if these psychologists ever studied basic engineering they would have made this conclusion a long time ago and also framed the discussion on stress differently.

As engineers we are taught that in the measurement of Stress (and Strain); stress is not the thing that is done to us by the outside world it is instead strain that the world puts onto us. Stress is the reaction to the strain.

Strain is what is applied in your life, stress is the reaction to that strain

During the deformation of sheet metal we induce a change in the material, it was one size and shape and we make it take on a new shape. The stress response in the metal increases the materials measurable resistance to further deformation. The stress increase helps the material prevent any excessive deformation by making local areas stronger which pass the strain onto the neighboring un-strained (or unstressed) areas. Materials that exhibit great formability are those materials that increase their STRESS in such a way to pass the applied strain onto other areas.

So when people talk about being able to change the effect of their STRESSFUL situations just by changing their outlook on stress, and that seems to not jive; after all the job is still difficult, it pushes us to the edge, it challenges us. It seems amazing that we could greatly affect the stress just by thinking about it differently.

But to this engineer I recognize that your job, or whatever situation, you think is stressing you out is not the STRESS. It is the STRAIN, and the stress is how I react to that given strain. Stress and Strength are interchangeable words in our field for a reason.

Flat pieces of sheet metal are only made into beautiful, stylish, strong, and safe products if we stress them beyond their previous strength capability. And that after doing so, the part is typically stronger for having done so.

Feasibility: What's in a word?

noreply@blogger.com (eric kam) — Mon, 30 Nov 2015 19:30:00 +0000

The prediction of part feasibility is a loaded question. For certainly the answer will always have to be, "It depends", or "Yes and no".

Capable processes are a subset of processes that may be manufacturable, which are a subset of the feasible processes

In a recent post on the AutoForm Engineering Blog, they shared a case study comparing a single part with two highly different proposed manufacturing processes. The part was a tube hydro-formed part which was initially proposed as a HPH, high pressure hydro-formed processed part.

In the field of Hydro-forming HPH processes are the more common and typically lower cost of the possible process selections. The alternative process "Multi-pressure sequenced" hydro-forming or "low pressure hydro-forming" (LPH) is often overlooked as it has been a boutique process offered by fewer/limited suppliers due to rights limitations for the previously proprietary process. It is often slightly more expensive to source and therefore assumed to be less cost effective.

However, when taking a more holistic look at the process and it's potential to produce safe and repeatable parts, AutoForm was able to demonstrate that it is likely that any increased initial investment in the alternative process could be proven to be well worth it, when looking at the big picture; Cost-Quality-Lead Time-Function.

Considering the big picture requires evaluating across more than one dimension.

The map is not the territory

noreply@blogger.com (eric kam) — Wed, 28 Oct 2015 20:00:00 +0000

In the definition of Computer Aided Engineering and Finite Element Analysis of sheet metal stamping it is necessary that we use numerical approximations of many physical behaviors. We represent curved surfaces of the tools and the sheets with small straight line elements connected at nodes (mesh).

Complex part geometry modeled as mesh

tensile test

We model the materials in-plane behavior using using data collected using uni-axial tension testing and assuming that it can correlate to bi-axial stretching or compression behavior.

approximated hardening curve

There are many who would argue that such a model should be replaced or augmented with a kind of Bi-axial Bulge test result, but will again have to default to some manner of numerical constitutive approximation of behavior in order to use the test data.

biaxial yield loci

In fact all computer modeling is predicated on the notion that a physical behavior is being modeled numerically. That many of the numbers one could use might be poorly correlated to reality, is an ongoing challenge faced by metal forming engineers everyday.

Friction is another one of those numbers, industry often uses a solitary coulomb friction value as the basis for modeling the highly complex "tribological" system that is the interface between the coated/uncoated piece of sheet metal, the lubricant or mill oil on the sheet, and the coated/uncoated surface of the stamping tool.

Metal forming simulation software model tools with perfectly rigid tools, traveling straight down and up, with perfect alignment, uniform and homogeneous pieces of sheet metal, that do not decay or fluctuate over time.

We often model drawbeads as "numerical" boundary conditions rather than modeling the actual deformation of the sheet through the drawbead shape. There is increasing pressure by users and their management to model everything "physically" instead of using enforced or applied boundary conditions--under the belief that this makes the model more accurate.

Why can't simulations just include all variables as models

With this insight people often want to know if there are systems that can couple a fluid dynamic tribological simulation, with a multi-physics tooling, operated by multi-physics modeled stamping presses, deforming blanks that perfectly model every potential degree of freedom and behavior modeled. These people are frequently disappointed that computer modeling can only go so far.

The response that must be given to these inquiries is a parable.

There once was a Emperor of a land known for its science,technology, engineering, and mathematics discipline. The Emperor charged his court cartographers with creating an exact map of the kingdom. When he reviewed their first attempt he was disappointed that the map was only a 2-dimensional rendering on velum. He charged them to do better. Their next map was a scaled model of the kingdom rendered at a scale to fit on a tabletop, but he was disappointed that on the map he could not distinguish the details of his palace gardens. The next version was scaled to fit in a park the size of a city, where he could see all the buildings but could not represent the energy of the market square or the vitality of the people in the city. The final map was a 1:1 model of the empire down to the tile, brick, person, and emperor and court cartographers themselves; however there were no longer any residents of the empire as they had all been moved to the map. The empire, without its inhabitants, its livestock, or even most royal leader fell into disrepair and was forgotten forever. The map once accurate, was now a perfectly inaccurate rendering of the place it had once been, but was no more.

Village with a Model village in the model model village in the model model model village.

The more accurate the map, the more it resembles the territory. The most accurate map possible would be the territory, and thus would be perfectly accurate and perfectly useless. - Neil Gaiman, Fragile Things

The most perfect simulation would be the tool, and the tool would be the simulation. It would be perfectly accurate and perfectly useless. In simulation we aim to model as accurately as we need to inform the necessary decisions that are better made before the tool. Perfect matching of the simulation to the reality of the tool is often impossible. What is more important is will the design of the tool, in the variable situation of the world produce a result which might not be perfect, but is acceptable. That is the question.

The use of numerical parameters in an attempt to get a perfect representation of reality may always fall short. Anybody who shows me a simulation with "perfect" predictions for strains, stresses, or springback kind of frightens me. It should not happen, EVER. Instead, what I find most informative about using simulation is the ability to recognize the trends in the behavior, increasing or decreasing beads make the thinning increase or decrease by what proportion. Modifications to the tool forces or flange timing results in what kind of springback differences. This is the most valuable thing to learn with simulation.

October 2015- Stamping and Dies Techgroup Web meeting- FABTECH preview

noreply@blogger.com (eric kam) — Wed, 14 Oct 2015 19:27:00 +0000

We hosted our 3rd Annual FABTECH preview during our October Stamping and Dies Technical Group meeting. With 3 special guests from the ranks of the Fabtech Exhibitors.

Paul Tamlin from Mayfran International
Chris Fletcher from Tower Oil
Tim Tuttle from Oriimec

They were kind enough to spend some time highlighting their companies offerings that they will feature at FABTECH2015.

Here is the recording of the meeting, with apologies for the amateurish nature of the recording.
Or the other recording from the WEBEX servers SME WEBEX server recording

FABTECH 2015 Preview- SME stamping and dies technical group

noreply@blogger.com (eric kam) — Mon, 28 Sep 2015 17:18:00 +0000

Each year before the FABTECH show we offer our members a unique opportunity to promote their activities at FABTECH Expo. This way they can be sure to maximize the value of their investment to attend the show and make sure they attract the most visitors.

If you would like to present at this years FABTECH preview please comment here and we will get back in touch with you to share how to be involved. If you just want to attend the preview just log into the SME WEBEX page that day and prepare to learn what is new and improved for this years Expo.

Meeting date Wednesday October 14th
Time: 1:00 PM Eastern Time (via SME webex)
5-10 minutes to showcase your technology
Introduce your products and the problem that it solves
Plug your booth and booth location at FABTECH

************ Edit, some early adopters:

Mayfran International, Paul Tamlin
Oriimec, Tim Tuttle
Tower Oil,Steve Lowery

Press speed and metal forming

noreply@blogger.com (eric kam) — Tue, 22 Sep 2015 10:00:00 +0000

There has been a fair amount of buzz and will likely continue regarding the "revolutionary" advances of servo drive presses. It is not likely to stop any time soon and leads to a fair amount of confusion regarding the benefits relevant to formability.

Let's take a stab at addressing this:

The part makes when we cycled the press slow, but when we speed up the press it failed. Can simulation predict this?

yes and no.
yes, simulation can be set up to recognize the effects of speed on the performance of material strain rate, lubrication speed dependency, and other effects
no, you might be disappointed to find that many early engineering simulations will not use strain rate dependent inputs (we have trouble getting the minimum information needed let alone the advanced)
no, you might also find that when you use a strain rate dependent material that the performance for splitting failure improves. See the figure below and note the steeper curve in the area just past Yield. This means that the material work hardens more effectively when deformed fast instead of slow. Stretching fast increases the formability

Strain rate adjusted tensile test results, note the increased slope from slow to fast. Better stretch distribution source: caeai.com

Why? It is likely that some other characteristic of the forming system is much more speed dependent. Performance of pressure systems (especially self contained in-die systems) might perform better slow than fast--achieving better balanced loading, preventing abnormal pressure build, and better tooling dynamics

The strain rates are not so different when you double the press speed as what is tracked for strain rate dependency. In the figure below note the strain rates. They are shown as degrees of magnitude difference from 0.0%/sec to 10%/sec and 100%/sec. However for forming in most presses if you change the cycle speed, this does not equal the same shifts as shown below. Increasing stroke rate from 10 strokes per minute to 40 strokes per minute might only be a strain rate change from 10%/sec to 30%/sec depending on where in the stroke the deformation happens. Stroke rate increases and proportional forming velocity and strain are dependent on the general sinusoidal relationship of press rate to ram speed (which is 0.0mm/sec at the bottom of stoke where most forming happens).

Forming at the rates of 1/s and 0.001/ sec at room temperature is very similar in behavior even though speeds of 1000 times faster are assumed

If that is so, what can we use simulation to predict?

If we find through simulation that the process is prone to major differences in performance when small changes to lubrication, pressure (pad or binder forces), distribution of that pressure, or minor changes in bead strengths are used. If the part is seen to oscillate between safety and failing conditions when only minor changes in the process parameters are used, it is certain that the process as engineered will be sensitive to the types of changes that occur in reality, but are not part of our standard simulation assumptions.

At this time it is not practical to create a simulation environment where we allow the computer to try to model all the intricacies of the stamping environment (flexible tools, in a slightly non-parallel alignment, on flexible beds of flexible presses, with fully visco-elastic behavior modeled friction at the interface of sheet to tools, etc).

But we can use robustness analysis to see if the forming process is sensitive to changes in the related inputs that we use.

	Engineering	Production
Pad/Binder pressure	Constant	Self contained? Manifold? N/C? not necessarily constant....
Lubrication	Constant friction number (coulomb friction)	Pressure and temperature dependent, sometimes non-uniform
Die/Press alignment	Perfectly parallel	In spec (0.001 in/foot)
Tool geometry	Ideal CAD/CAM/FEA	We are all human…
Sheet metal properties	Single set of inputs, based on constitutive models...	changing day-to-day, run-to-run, high/low, and everything in between

Simulacra and Semantics: 3-D draw beads vs Line beads

noreply@blogger.com (eric kam) — Thu, 03 Sep 2015 11:00:00 +0000

The more accurate the map, the more it resembles the territory. The most accurate map possible would be the territory, and thus would be perfectly accurate and perfectly useless. - Neil Gaiman Fragile Things

Starting off a blog post on metal forming simulation with a quote from fantasy fiction author and graphic novelist Neil Gaiman might seem a little odd, but stick with me. I might be able to pull it all together.

The larger concept and initial dictum, embodied in the quote above, is attributed to the German scholar and the father of general semantics Alfred Korzybski.The semantic dictum that the map is not the territory goes to the root of simulation technology today and in the future. How much of the model must be modeled exactly and what can be simulated using proxies to ensure that the model is valid and useful.

When I first started in metal forming simulation as a field (a scant 16 years ago) we were hobbled by computer technology and limitations with finite element analysis methodology. To simulate the draw forming of sheet metal body panels we needed to use many simplified boundary conditions just to get simulations to run.

limited simulations to draw only for splitting and thinning
moderately coarse meshes
settled for dynamic explicit models rather than implicit models
simulated draw beads with nodal application of external forces
modeled complex sheets as modified membranes rather than shells or solids

drawbead model from http://www.sciencedirect.com/science/article/pii/S0020768308000668

Line bead models and other simplifications of boundary conditions were essential to the modeling of those early analyses. Without them even the most basic of forming operations would have not been possible with even last decades computers.

As computers technology grew with faster processors we could loosen some of those constraints. As memory became more available we could expand the use of implicit solvers rather than "tricking explicit solvers" to run slower deformations. We could make the meshes more dense, we could expand the use of elastic plastic shells instead of modified membranes. We could look beyond draw operations only and review multi-stage processes and perform springback predictions. We could make the meshes more dense, we could expand the use of elastic plastic shells instead of modified membranes. We could use more dense meshes and we could even model the formation of draw beads.

The modeling of drawbeads is one area where recent evolution in computer and FEA technology have made many assume that they get better results by modeling the full forming of the draw bead and use the draw bead throughout the forming process. This often results in a complete offsetting of recent gains in computational time--simulation that took 12 hours to run 6 years ago, can now run in 1 hour with the use of line beads. Replacing the line beads with 3-D bead geometry can increase computation time to 6-10 hours again (CPU hours, using more CPUs can still post results in reasonable real time, but at greater costs).

Many have rationalized that this increased cost improves the overall accuracy of the simulation. Setting the bead creates deformations locally in the sheet which can propagate over wider areas. These "breakdown buckles" can often significantly change the behavior of the die. Also, the 3-D beads use material in their forming. This often meant that results of previous line bead simulations had to be adjusted manually to account for the additional use of material in determining the blank size. These are real and valid reasons for wanting to simulate the forming of the bead.

However, modeling the metal sliding through the draw bead continually for every time-step in the analysis is a repetitive and limited value computation. For the most part once we understand the distortion caused by the bead setting and the amount of material expended in making the bead shape, the effect of the themselves will be fairly consistent over the entire forming operation. There is little added value if the material that draws through the bead is trimmed off as engineered scrap. If the material that moves through the drawbead is left on the part, then the deformation of the material through the bead is essential, and needs to be modeled for final validation, but in trying to identify which drawbead forces are needed, it is stil much better to iterate with the line bead.

A new feature in AutoForm Engineering AutoForm-Solver allows the user to define a draw bead as a 2-D profile with geometric parameters, the software simulates the binder setting with the 3-D shape as defined, then removes the bead shape from the mesh. The bead will instead continue to act on the material interior to the bead through the application of adaptive line bead restraint and reaction forces. This reduces the computational costs for the remainder of the forming to stay relatively lowb but with a majority of the benefit.And we could discuss the many ways that having the additional geometry might in fact lead to worse results that have numerical aberrations or inaccuracies.

AutoForm Engineering GmbH blog on drawbeads. link

Yes, the most ACCURATE model of the process would have all the elements in it all the time, but then its value as a simulation is lost. It takes too long or is too complex to use.

The MAP is not the Territory, and the Territory is not the MAP

The simulation is not the tool, and the tool is not the simulation

August 2015, SME stamping and dies techgroup meeting: SPRINGBACK

noreply@blogger.com (eric kam) — Tue, 18 Aug 2015 18:51:00 +0000

We hosted our August meeting of the Stamping and Die Technical Group and Metal Forming simulation technical groups with a discussion on Springback. The presentation was made by Eric Kam of AutoForm Engineering GmbH.

Basic definition of Springback

We are hoping the recording is available soon, but in the meantime the slides used are mounted online at: SME Google Docs

The assumed behavior of springback in order to try to "compensate" if the net shape in blue is the target, and the red is the response, then compensate by forming to the dashed line. (not always feasible to achieve).

During the meeting we discussed how people are modeling springback and what they try to do with the results from simulation. Namely Springback compensation and springback minimization.

An indirect springback response common in stretch flanges, the area adjacent to the flange is distorted

One challenge of springback control in the real world is that with minor differences in forming conditions, significant changes in springback response is likely.

in the end we discussed how springback modeling is not as simple as predicting "safe or failing" as any difference in prediction from expected is different and therefore wrong. You cannot over engineer for springback. It is essential that when trying to address springback that one recognize the springback is a variation issue, and that a robustness assessment is needed to know if the springback that you predict is repeatable and stable, lest you try to compensate for a response that is less likely to happen in production.

Some additional background reading:

Rapid blank estimations from inverse one-step FEA

noreply@blogger.com (eric kam) — Thu, 18 Jun 2015 23:09:00 +0000

One of the simplest tools to employ in digital process planning for sheet metal stamping is the inverse One-step finite element analysis method. Very common to use during initial part design as a very early geometry based form check, inverse one-step FEA is widely distributed as part of sheet metal engineering software productst

Section view part and "Minimum blank" from inverse one-step

One of the reasons for its widespread use is its simplicity of operation and set-up. In order to run a one-step analysis of a sheet metal part the only required inputs are the part geometry and the material properties. Given that information, just about anybody can kick off an analysis. The part shape is "meshed" and the material mechanical properties and thickness applied to the mesh. The mesh is flattened virtually and the the amount of stretching and compression (strain) is computed for each mesh element. This flattened state is the assumed initial state of the part or blank.

The amount of deformation predicted in the inverse forming is them applied to the part geometry as a reasonable prediction of expected deformation as a blank is formed into the shape of the imported part. Thus the name inverse one-step. The prediction is based on an inversion of the intended forming process and the forming is computed in a single step.

The meshed blank and meshed part

Because the input requirements are so low, these solutions are promoted very actively to product designers and cost estimators as a way to predict feasibility without waiting for manufacturing input. Additionally, the blanks are available much earlier circumventing any need to wait for complete die process planning and die engineering to predict a possible material cost requirement.

Part shape and blank prediction from Inverse OneStep

One of the major pitfalls with using the minimum blank which is estimated from the simplest of inverse one-step methods is that the blank prediction is merely the same volume of material which makes up the finished part. This might be acceptable if the part is effectively "crashformed" using a blank which is fully developed. This is the case for some parts but certainly not all.

One method to address this short coming is to apply some expansion to the initial blank prediction to serve as "addendum" material. This would require the user to make some editing changes to the curve to represent the additional material needed. The problem again is that the user must somehow guess how much material is needed and be aware that they could easily over estimate the needed material. Also the existence of any additional material will affect the material flow, and alter the prediction of part feasibility.

The animation above illustrates the end effect of a blank estimate from part only one-step being fed into a die which ultimately needed a drawn blank. It also illustrates how using assumed binder and addendum as an input to the geometry based formcheck can result in a better blank.

Possible draw configuration for the part, run double attached with developed flanges

In order to get a better blank prediction it would be necessary to use, instead of the part shape only, a potential shape of the draw tool which could be used. If available nearly any one-step FEA code should be able to use the "draw development" as the initial input and predict the blank needed to form that shape. The challenge is the availability of the 3D CAD surface which represents the possible tool configuration.

Because the aim was to predict formability prior to starting the tool design in earnest it makes little sense to require a full 3D tooling design to arrive a blank prediction or estimated material costs. For this reason development of software features which can automatically propose a plausible tooling geometry for draw which can serve as the input for the more detailed one-step analysis.

Section cut through the part with attachment, addendum, and wrap surfaces added with the blank

While the 3D tooling proposal will likely not be what well trained tooling engineers might create, it is a reasonable tooling requirement for potential material requirements. In most cases while the shapes created are sub-optimal the material requirements are a much more reasonable starting condition rather than those blanks shapes based on part shape only.

Section through part, attachement, addendum, and wrap surface with the blank estimate

Formability plot with issues predicted under "optimal" stretching for draw, maximize stretching over majority of part

The blank prediction from the automatically developed draw tool accommodates not only that draw addendum will be needed and its requirements for additional material will be based on the amount of stretching anticipated and relative depths, but also recognizes that any features not formed complete in the draw (such as backdraft geometry) will need to be developed out onto available draw surfaces and material needs to be provided for.

Blank nesting for single part without addendum

Blank nesting for double attached blank

Images courtesy of AutoForm Engineering GmbH

using AutoForm-StampingAdviser^plus

and AutoForm-ProcessPlanner^plus

June SME Stamping and Dies Technical Group meeting: Diversity in manufacturing and engineering

noreply@blogger.com (eric kam) — Thu, 18 Jun 2015 17:51:00 +0000

Brianne hosted a great discussion earlier this week when she invited several of her colleagues in engineering and manufacturing to share their insights into diversity in engineering and manufacturing. They all came to the discussion with different perspectives as Women in Engineering.

Brianne was born into it and followed here dad's footsteps into the family owned machining and tool & die business
Jill is a former school teacher who has found more growth in her career in engineering and manufacturing than the state of Ohio allowed
Julia is a recent graduate of engineering program, who transferred out of a health sciences curriculum into engineering

The video is embedded below.
Also an alternative video stream is here hosted by WEBEX: Play recording

Faster Horses... Virtual Tryout?

noreply@blogger.com (eric kam) — Mon, 08 Jun 2015 22:21:00 +0000

According to online legend, Henry Ford is attributed with the quote, "If I had asked the customer, they'd have asked for faster horses." This quote whether real or not, is used to point to the fact that often when innovating one must go their own direction, and often point the customer to something that they did not realize they needed or wanted. But also, it can point to how when trying to address customers actual needs, one might need to shift the paradigm. Personal transport used to mean horses, prior to the launch of the automobile there was no market for other forms of powered personal transport.

This is the first thing that popped into my mind when reading the blog at AutoForm.com. It posed the question:

Just how proud should one be of the virtual tryout title?

Source: https://www.flickr.com/photos/triplea4/ under creative commons 2.0 remix and attribution

The author contends that after 20+ years of simulation integration with the metal forming industry, people are still talking about how using simulation can "reduce tryout time" and "eliminate the need for soft tool". But most of those time savings are already factored into the now greatly reduce product development life cycle. Back when simulation was new and shiny, vehicle development programs were often 48-60 months from concept to production release. Now vehicle development plans are only 24 months from end of styling to production.

Infographic: Soft tool development, virtual tryout, and process engineering

Proof tooling is so antiquated that recently I had to describe how it used to work to a person who has worked in stamping for last 15 years. They had never heard of proof tools, they only know a world where stamping simulation is a sunk cost. It just must exist. When this person asked how can one simulation product promise to eliminate an engineering step that does not exist, what good is it?

The tryout press and proof tool have been replaced with a virtual press and virtual die tryout. Is this really as good as it gets? Faster design and computation of stamping process is little more than guessing faster—well- informed, educated guessing based on know-how and experience, but still trial and error, still guessing.

When I worked in the stamping plants chasing production improvements, we essentially defaulted back into the guess, try, check, repeat mode of tooling development. There was no better way. If I could load one piece of technology into the way-back-machine for my younger self, I would want AutoForm-Sigma containing a lifetimes worth of trial and error in a single file which I could use to "look up" the effect, would have been a tremendous time and effort saver.

Simulation: Draw beads difference between simulation and reality

noreply@blogger.com (eric kam) — Thu, 21 May 2015 12:25:00 +0000

In my time as a metal forming simulation applications engineer. The frequency of calls from users expressing dismay that the stamping process that they had engineered had shown one result in simulation, and when the die was completed a different result. Often they might say "the simulation was wrong" or something to that effect. Almost every time we could get to the bottom of the differences when I would ask them, "How is the binder spotted?"

Draw dies will often employ binder (or pad) force in addition to draw beads to restrain the metal flow into a die and affect a stretch in the material. In the illustration below, the binder force is being applied through the use of a press air cushion. The beads are shapes in the binder which when stamped into the flat blank will serve as a restriction to free material flow into the die cavity.

simple schematic of a draw die

Metal forming simulation is very often concentrated on identifying what shape of bead (restraining effect) is needed to successfully for the part--no splits, wrinkles, good shape fixation, and suitable surface quality. Stamping process engineers and layout designers will try various die shapes, blank properties, bead restraint, and applied binder force until they reach the safe combination. Once a safe combination is found they will release the design of the die face and beads to be designed into a tool and ultimately cast and machined into a stamping die.

Drawn hood panel with draw bead geometry

Once the die is complete, die tryout will validate the tooling build and stamping set-up. This is where those support calls come in. Very often the first few hits with the new draw die may show significant differences to the simulation, which leave the engineers questioning the validity of the simulations. But often the differences are not with any aspect of simulation accuracy but with the assumptions made by the engineer and the set-up of the tooling for simulation.

Binder fully spotted, binder force transmits directly to the sheet

Most often in simulation the user will specify a binder force. In simulation the binder force is a sheet boundary condition. Unlike a physical die, the force is transmitted directly to the sheet rather than the tool. It is as if the die can be assumed to have been spotted perfectly to equally transmit force to the exposed surface of the sheet. Every exposed surface of the sheet will receive some transmitted force from the binder. In the event that the blank shape is different or the blank is located on the binder slightly off center, this will result in binder forces on the sheet that can be very different than what was assumed for the simulation.

Force transmitted to the sheet = Binder Force - Bead Uplift Force

When the restraint of the material needs to be changed the applied force, bead shape, and addition of shims to equalizers/spacers can all be used to affect metal flow. However, while this may seem to give the tool makers many opportunities to adjust metal flow, it can lead to unstable processes by leaving open too many adjustment points that might fluctuate with production noise.

Binder spotted, but with equalizers maintaining gap > t

However, it does occur that the tool might employ equalizer blocks. These equalizers (or standoffs, kiss blocks, spacers, shims) if used can maintain a gap between the upper and lower tools. When the gap is maintained at greater than sheet thickness, the sheet never gets fully clamped between the upper and lower die. No force is imposed to the sheet, all metal flow restraint is a function only of the bead shape. If the metal flow needs to be adjusted, the only change is the bead shape itself. This often seems challenging to the tryout source, as adjusting bead shape required grinding and/or welding of the bead to affect its shape. As contact is not a metal flow driver, draw dies set like this are inherently insensitive to changes in blank shape and location.

Sometimes toolmakers might also, reduce the height of the equalizers to less than t. When this occurs clamping contact between upper and lower tool is reintroduced as a potential control. Seemingly addressing the problem short term, but introducing a potential source for process variation. Sometimes tool and die makers might add shims to the equalizers. While this will not change any clamping effects (assuming the gap was already greater than t) it does make the geometry of the bead effectively change. Adding shims to the figure above would make the bead seem shorter and change the angles of contact between the upper and lower bead shapes.

Binder partially spotted, force transmitted to localized area of sheet

In other cases the spotting of the die may be unevenly distributed: either by design or accidental. If the spotting is not uniform then the force applied by the sheet will instead be applied only to the areas where the sheet is clamped between the upper and lower tools. It is common for some products that the tool makers will intentionally clear areas outside the draw bead creating a partial spotting only inside the bead. This makes the stamping process somewhat less sensitive to blank size, shape, and location variations.

This happens accidentally on occasion as a result of incomplete die finishing, castings that flex locally, poor machining tolerances, cutter marks, galling or damage to the surface of the die. Any localized areas with poor spotting will be clamped with higher proportions of the transmitted binder force, while areas where the poor spotting result in increased local gaps, there will be less material restriction and more material draw-in.

For these reasons, it has always been one of my first areas of inquiry when faced with troubleshooting differences between the "as-built" tool and the "as-engineered" conditions predicted by simulation. If you see differences in your tools:

verify that spotting is uniform over the entire surface of the binder
verify that spotting is used as intended (did simulation apply force through the binder or run "gap controlled")
verify that equalizers are sized to achieve the intended conditions
verify that the blank is locating itself as planned in the simulation

Once these conditions were met, we find that correlation to simulation is very good (if the material is realistically matched).

NIST Roadmapping for Innovation: Sheet Metal Forming

noreply@blogger.com (eric kam) — Tue, 05 May 2015 15:18:00 +0000

After a fairly lengthy amount of discussion with a group of academic (mostly), industry research, and industrial consultants and practitioners it appears that much of stamping innovation is challenged by either incomplete, in accessible, or incorrect material characterization. Or subjectively that is what my take away was.

After breaking into 5 groups who were defined by Materials Ferrous, Materials Non-Ferrous, machinery-equipment-lube, Simulation and ....... (one other that escapes me now) we were tasked to as groups identify each as a group our "top 5" challenges, barriers, opportunities. As with any brainstorming activity suggestions were quite widely divergent among the members, but in the activity of ranking and prioritization any patterns of consistent identification will come out. The consensus for the group I sat it and for most of the other groups was that consistent and reliable characterization of the materials we engineer and use was uniformly an issue.

Some of the discussion went further afield by saying that advancements in strain-path, strain-rate, stress-strain under biaxial loading, micro-structure, crystallography, phase distribution and other measurements needed to somehow be modeled. I was in this discussion thwarted when I pointed out that we already have challenge of making use of the data that we have from tensile testing as it is.

Yes, tensile tests are not as good as Bulge testing for capturing the full range of biaxial material behaviors. But even with the relatively simple data that can be collected from uniaxial Tension tests, we still fail to make readily available and accessible, actionable data on the behavior of our materials. How will this be made better by trying to promote a standard of testing that is more "abstract" and at this time has very little infrastructure for widespread use?

went a little something like this:

ME: Today we could improve many simulations if we stopped applying ""arbitrary" min/max values to extrapolate flow/hardening curves. Instead, use readily available tensile raw data to model curve, and 3 directions to capture yield surface. Improves error in simulation given by mis-use of "power law".

Them: but tensile test is not good only Biaxial Bulge is acceptable.

ME: how do you propose people use the bulge test data?

Them: they can use the n, YS, TS and R we derive/estimate from the bulge test as entry into the flow curve model in FEA. Using a "power law" equation

ME: but does that not introduce the same kind of derivation/extrapolation error i suggested we could eliminate if people used the raw Tensile data?

Them: Tensile data is no good...(rinse-repeat)

NADDRG: May 5th 2015 meeting agenda

noreply@blogger.com (eric kam) — Thu, 30 Apr 2015 12:10:00 +0000

The NADDRG Symposium is scheduled for May 5th

Northwestern University

Rebecca Crown Center, Hardin Hall

633 Clark St. Evanston, IL 60208

If you have not already paid, please plan to pay at the door (check, credit card, or cash).

The meeting follows a workshop held on Monday:

The Roadmapping workshop is to chart a development and research needs/path for innovation and improvement of the sheet metal forming industry. I will attend the workshop and welcome any input from the readers here to bring along as "proxy" input for the roadmapping exercise.

The tentative schedule for the meeting Tuesday May 4th is:

8:30 sign-in/registration
8:50 announcements and introduction
9:00-12:00 Presentations
11:15-12:00 Working lunch
11:50-2:00 Lab tours:

NUANCE
EA Forming
CLAMMP/Tension
Incremental Forming,
Laser
Microrolling

2:00- ~5:00 Presentations
~5:00pm End meeting

The scheduled talks are shown below. The order of the talks has not been finalized.

Dr. Jian Cao

Northwestern University
Roadmapping Workshop Summary

Peter J. Ulintz

Anchor Manfg. Group Inc.
Eco Pickled Surfaces (a.k.a. Slurry Blasting) - Some Recent Issues in the Press Shop

Dr. Bernard Levy

BS Levy Consulting
Shear Edge Stretching

Dr. Hyunok Kim

EWI Forming Center
A modified FLD to predict the edge cracking with AHSS and Aluminum materials

Cliff Butcher

University of Waterloo
Edge Formability of HSS

Dr. Yannis Korkolis

University of New Hampshire
Hole expansion of anisotropic Al-6022-T4 sheets

Dr. Xin Wu

Wayne State University
Constitutive Law Based on Thermodynamics of Microstructure

Son Pham

NIST
Crystal Plasticity-based Constitutive Modeling for Complex Forming Paths

Dr. Brad Kinsey

University of New Hampshire
Numerical and Experimental Evaluation of the Onset of Necking during Biaxial Testing of 1018 Steel

Youngung Jeong

NIST
Forming limit diagram prediction using a self-consistent crystal plasticity framework: experimental validation and application

Dr. Michael Worswick

University of Waterloo
Time dependent methods for FLD characterization applied to warm forming

Dr. Yuanli Bai

University of Central Florida
Ductile fracture prediction for metal sheets using all-strain-based modified Mohr-Coulomb Model

Dr. Thomas Stoughton

GM Global R&D
Center Why Does Industry Need An Accurate FLD?

Please see www.NADDRG.org to pay with a credit card and for the latest copy of our Bylaws. You'll also find instructions on how to apply for membership.

Great Designs in Steel, 2015: Come and meet us

noreply@blogger.com (eric kam) — Wed, 29 Apr 2015 14:07:00 +0000

In May I will be at the "Great Designs In Steel" conference in suburban Detroit. GDIS 2015 is presented by the SMDI (Steel Market Development Institute) and focuses on applications of steel and high strength steel products in particular.

Throughout the day there are a number of presentations and seminar sessions wher subject matter experts share what is latest and greatest in Vehicle Body-in-White design regarding the application of steel. Component suppliers, material suppliers, simulation and engineering providers will be in attendance and exhibiting.

I will be there with my employers (www.autoform.com) where we will be exhibiting the current version of our market leading Sheet Metal Process Engineering and Validation software. We welcome any participant to bring along their sheet steel product designs to be imported, process engineered, and simulated live in our software throughout the day*.

Challenges that can be addressed:

what is a reasonable tool cost for conventional stamped parts?
is the part manufacturable?
will the process produce repeatable parts?
if the part is hot-stamped how strong will it be in the end?
will tailor welded parts yield desired results?
how much springback can i expect?
how can I resolve the springback?
and many more.....

Registration: Link (http://www.cvent.com/d/d4qyl9)

Source: WardsAuto.com

* understand that this is a public event, please don't bring anything you can't have the public viewing

Simulation and reality mismatches

noreply@blogger.com (eric kam) — Tue, 28 Apr 2015 15:22:00 +0000

My employers blog recently featured post on Tips-and-Techniques for stamping simulation. The post proposed that among things to check when attempting to correlate the simulation to reality draw-in is one key place to begin.

Sheet metal properties will vary from batch to batch (sometimes within a batch)

The ugly truth is that with your incoming properties of the material, the tool, press, lube, etc are all very prone to variation over the life of the product. This results in changes to our outputs (your production).

We spend a great deal of time and energy trying to ACCURATELY model the deformation of sheet metal. But when the inputs do not reflect a plausible reality or a limited sub-set of reality; the accuracy of the model is questionable. GIGO (Garbage In Garbage Out).

Without even delving into the reliability of Constitutive models for mechanical properties, but only at the availability of ANY information, there is much to be done in improving the information availability.