Incremental Forming-Simulation Software
Back in June 1987, I was tasked with my first formability assignment as a second-year apprentice at the famed UAW/GM Die Engineering Apprenticeship Program. We created die layouts on Mylar by using descriptive geometry with dividers, triangles and pencils to develop the die face for draw dies. The punch opening and binder surface was just a line drawn by hand—old school.
The two-dimensional die layout was the blueprint to build a three-dimensional model of the die face engineering out of plaster. We executed engineering changes and what-if analyses with modeling clay and a sculpting knife, and machined tools from the plaster models. Formability analysis occurred for the first time at the tryout press.
The automotive stamping world changed when the workhorse SAE 1008-1010 low-carbon-steel family was tossed aside in favor of high-strength low-alloy (HSLA), dual-phase (DP) steels and other materials with tensile strengths exceeding 1000 MPa—I refer to these as super steels—for structural stampings.
Not long ago, wire EDM revolutionized the manufacture of die sections, replacing a generation of thinking that the only way to machine steels was to mill and grind sections. Now there is another revolution underway—software technology that accurately predicts the formability of sheetmetal stampings before any steel is ordered, let alone sent to the wire-EDM machine. This no longer is your daddy’s die shop.
While automotive product geometry has not changed much in the last 25 years, stampings now are made from materials with a fraction of the formability characteristics of materials used in the past, all while tooling budgets seemingly have shrunk and continue to do so. The margin for error is nearly zero, and the stamping-die business is a high-risk high-wire act.
Your safety net: Formability-simulation software and the ability to predict production sheetmetal formability issues.
One Step vs. Incremental Simulation
I often hear from die shops, “Oh yeah, we did a simulation on that.” Actually, they did what is known as a one-step simulation, a relatively fast directional indicator of potential forming issues. The process serves as a compass for early feasibility to a product engineer—a mere substitute for not having an experienced die engineer review the design early in the development process.
In addition to a one-step simulation, shops can opt for a multi-step simulation—a series of one-step simulations that enables an engineer to segregate drawn features from flanged features, such as return flanges. At the die shop, you need a map—not a compass—to navigate your way from quote to buyoff.
Incremental forming-simulation software is as close to standing at the tryout press as you can get. This simulation technology digitally replicates the actual forming process. It requires form-tool surfaces and mechanics, and blank geometry. The software uses the tools and blank to calculate formability throughout the sheet thickness during each fractional increment of press stroke.
Software PlatformsIncremental forming-simulation software platforms are built around two types of users: the FEA expert and your average everyday die guy. The FEA platforms require an expert to create what is known as a mesh; mesh size helps to determine simulation accuracy. The software still requires engineered form-tool geometry and a blank. Meanwhile, software platforms designed for the average die guy automate meshing tasks.
So, how do you know which platform is right for you? Those that actually know what the phrase “transversal shear stress” means and can use it in a sentence should typically opt for the FEA platform. Otherwise, I recommend using software that automates the meshing procedure.
The software brand choice may already be decided for you depending upon your customer base. Some companies—VW and BMW, for example—govern the software brand for their incremental forming simulations.
Keep in mind that most FEA experts can successfully use any of the software brands on the market, provided that the die processing is performed in CAD by a die guy. The converse is typically not true: Most die guys struggle to use the FEA-driven simulation brands because their expertise is in die engineering and not FEA.
Some simulation software offers built-in parametric die face engineering (DFE) functionality. This enables the modeling of the form-tool architecture for each forming operation from the imported product CAD file. The parametric DFE rapidly yields a production-intent die process. Used for production feasibility, this tool-modeling method proves useful at the quote stage of the die-design-and-build transaction. Parametric modeling takes the DFE process from hours down to minutes, while providing the ability to reliably test various product change proposals or alternate die processes without ever having to launch CAD software.
Once the project has been awarded and the stamping has been processed into a progressive-strip layout or transfer flowchart, the forming surface design begins in CAD. This CAD-driven DFE then is imported into the incremental forming-simulation software for analysis, to validate—or in some cases invalidate—the form-tool surfaces and mechanics.
There is nothing wrong with having an FEA specialist perform the simulations. If the specialist can create die processes with production-die-intent forming-tool surfaces complete with architectural mechanics, he can perform the simulation analysis on his own from start to finish.
However, if the FEA expert does not have the knowledge and experience to process the stamping and build the die-face engineering, a skilled die engineer will have to perform this work.
The Value of Live Demonstrations
There is no substitute for seeing software in action, via a live demonstration. Live demos have a way of cutting through smoke and mirrors, so schedule a meeting with at least the software platforms on your side of the staffing equation. And to help prevent buyer’s remorse, meet with software vendors from the other side of the staffing equation as well. The hour or two invested in the sales pitch at least will provide confirmation that your purchase decision was the correct one.
Be prepared for each platform to pitch their brand as the one having the most accurate solver in the business. This kind of statement might have had merit five or 10 years ago, but my independent benchmarking effectively takes accuracy out of the equation. And while all of the software brands are reasonably accurate, none are perfect. But don’t take my word for it—find out for yourself with live demonstrations.
Beware: Many software companies will perform a canned demo, performing the step-by-step simulation live before your eyes. Then, to avoid the wait time for results while the software runs, they show the finished results from a simulation run ahead of time.
My recommendation: Following this test-drive demo, hand the software salesman a USB drive that contains the tools, blank outline, material properties, pad travels and pad forces for a die you recently built. A combination of drawn features, stretch flanges and compression flanges on the stamping you provide covers the gamut of the more difficult forming issues for the software to predict. Tell them you want to not only watch the setup, but you want them to solve for final validation.
You already know the results of your stamping, in terms of splits and wrinkles. The simulation software should produce nearly the same results experienced at the tryout press. A good simulation engineer will be able to set up this sight-unseen job in about the time it takes to brew a pot of coffee. After you have met with all of the software brands that interest you, review the results of your test stamping and compare the accuracy for yourself.
Training and Tech Support
Training is essential. Incremental forming simulation is not an intuitive technology engineers can use without proper training. Don’t fool yourself into thinking simulation software is just like CAD. So, be sure to choose a software platform from a vendor who provides formal training for new users; ongoing training for experienced users is a plus.
Further, there will come a time when the simulation engineer will require technical support to solve an issue. The potential issues range from requiring a click-by-click guide for advanced functionality features not used every day, to needing a second opinion on particular simulation results that appear out of the ordinary. Having access to experienced engineers, knowledgeable with the software and who can provide technical support, is a huge benefit. Be sure the software platform selected has a solid technical-support staff that is available during your business hours.
Inhouse vs. Outsource Calculations Assumptions: $45,000 annual lease for incremental simulation software Brand A 120 unique part numbers annually that require simulation $7500 per part number for outsourced simulation services Breakeven Analysis: $45,000/$7500 = six part numbers Payback Forecast: (120 – 6) x $7500 = $855,000 net savings from avoided outsourcing fees $855,000/$45,000 = 19X net payback
Technical support also can come from a well-organized user community, often found with suppliers of widely used specialized software platforms like incremental forming-simulation software. These user communities are filled with smart engineers that live by the mantra “find a way, or make one.” They are prepared to offer tips, techniques and workarounds for everything. This translates into minimizing simulation engineering time by learning and sharing methods that work—and those that do not. Look for software platforms offering well-organized user-group events.
Inhouse vs. Outsource Calculations
$45,000 annual lease for incremental simulation software Brand A
120 unique part numbers annually that require simulation
$7500 per part number for outsourced simulation services
$45,000/$7500 = six part numbers
(120 – 6) x $7500 = $855,000 net savings from avoided outsourcing fees
$855,000/$45,000 = 19X net payback$45,000/120 = $375 inhouse software cost per simulation
The Value Proposition
Justifying an investment in simulation software requires a shop to project reductions in tryout time and the resulting cost savings, which can prove challenging. I propose that incremental formability-simulation software has become an absolute requirement today, as did investing in wire-EDM technology became 20 years ago. It is the cost of doing business. So, if we must perform simulations anyway, the decision to invest in inhouse simulation capabilities versus outsourcing the jobs depends on the volume of simulation work. Think annual volume, since most simulation-software platforms carry annual lease terms.
The math is simple (see the accompanying sidebar): Take the software pricetag and divide it by the cost of outsourcing the simulation for one die. The result is the number of simulated parts needed to attain an annual break-even point. If your actual quantity of dies is less than this number, then outsourcing makes sense, but if the number of dies exceeds than this number, investing in the software will pay off.
The payback analysis, or return on investment, depends on how many part numbers will require simulation in one year. Take the number of parts you plan to simulate in one year and multiply by the cost of outsourcing each simulation. Then take that number and subtract the cost of the software. This is the net savings; ROI is net savings divided by the annual cost of the software lease.
Knowing that using incremental forming-simulation technology is the cost of doing business today, any additional value gained by avoiding tryout issues or solving production problems is what financial folks call residual value. Us die guys just call that gravy. Whatever you decide to call it, the only value you can reliably quantify is the inhouse versus outsource decision.
Optimizing Simulation Accuracy
With a simulation platform selected, the job becomes ensuring that the results of the simulations replicate actual production results as closely as possible. Studies I have recently performed reveal that simulation results can consistently correlate to actual stampings within one percent, provided certain guidelines are followed when deploying the simulation software.
First, while every software platform comes loaded with database of material properties, these libraries are generic values for generic materials. If your goal is to predict stamping performance under worst-case production conditions with production sheetmetal, these generic material properties simply will not suffice. Ask your customer to define the specific material properties for its stampings by providing you with a stress/strain curve plus n- and r-values. If they are not able to or willing to do this, then run your simulations with material properties at worst case values. This equates to the highest yield strength and tensile strength, and lowest n- and r-values.
Run solver settings that provide the most accurate and refined calculations for production-stamping feasibility and validation. This roughly equates to using eight elements in each radius; having initial element sizes no longer than 20 mm; slicing the material thickness into 11 layers; and using relatively small time steps.
A simulation with correct die-face engineering, correct preprocessing and correct solver parameters is useless if the interpretation of the results is incorrect. The engineer must define the conditions or requirements that enable the project to continue or stop, by using specific pass/fail criteria.
Always begin with the forming-limit diagram (FLD) and the safety margin needed to account for real-world production variables not accounted for in the simulation. At Johnson Controls, the stampings I am responsible for must pass stringent government-mandated safety standards, so I use a 20-percent safety margin. Other industries may use a 10-percent safety margin.The safety margin defines pass/fail. If all elements fall below the FLD marginal limit line, the stamping is FLD safe—pass. If any element touches or lands above the marginal limit line, the stamping is not FLD safe—fail.
The formability review continues only if the stamping is FLD safe, by analyzing thinning and thickening at the die-cut features. Why just the die cut features? First, unless the product design has a maximum thinning requirement for, say, vehicle safety, then the FLD shows if the combination of forming mode (plane strain, draw or stretch) and the strain capability of the material is sufficient to produce the formed feature without splits. Thinning capability for a given material type and thickness varies with product geometry and forming mode.
Second, it is beyond the current capability of simulation technology to replicate the localized work hardening and fracture, due to cold-work die cutting at trimmed or punched features. Why is this important to the analysis of simulation results? The software cannot predict edge cracks at die-cut features. Sheetmetal can have tensile or compression edge cracks even with a passing FLD.
To solve this dilemma, we use maximum allowable thinning and thickening percentages at die-cut features to predict edge cracks. The percentage varies with material type and thickness. JCI has extensively researched and tested the family of material it uses in its products to determine the maximum allowable percentages to prevent edge cracks. You need to consider your own edge-crack conditions and develop the criteria needed to determine pass/fail.Do what works for you and do it consistently. Pass or fail. MF
See also: Johnson Controls Inc.
Related Enterprise Zones: Tool & Die
There are no comments posted at this time.