WebCAE

Pilot program

Send us one STEP assembly. We'll send back a full FEA study.

The WebCAE pilot is a simple, structured evaluation. You share one or two CAD assemblies, we prepare the model, mesh it, run the analyses you asked for, and write up what we found. It's free, it runs over email, and the report is yours to keep either way.

Happy to sign an NDA before any file is transferred. Typical turnaround is 5–10 working days.

Cost
Free during the pilot
Input
STEP (AP203 / AP214 / AP242)
Output
Report + live model
Turnaround
5–10 working days
Who it's for

Who this pilot is for

It's built for teams who already do structural analysis — or know they probably should — and would rather evaluate a solver on their own geometry than on someone else's demo part.

  • Mechanical and design engineers

    You have a part or assembly you're not fully sure about. You'd like stresses, deflections and a safety factor on the geometry you're actually working on, not on a textbook example.

  • Engineering managers

    You're wondering whether simulation can move earlier into the design loop without adding a seat of a heavyweight FEA package and the training that comes with it.

  • R&D managers

    You're evaluating tools for a new product line and want something you can compare fairly: our results against your existing solver, your benchmarks, your acceptance criteria.

  • Small and medium engineering companies

    Analysis is currently outsourced, done in a spreadsheet, or skipped when the schedule gets tight. The pilot gives you a grounded read on what in-house simulation would cover.

Problems

Typical problems we can evaluate

These are problem classes the solver handles end to end today. Each one maps to an analysis type in the table below.

  • Load-carrying brackets and frames

    Peak von Mises stress, safety factor against yield, and deflection under service load.

    Analysis type: Linear static

  • Welded and bolted assemblies

    How load transfers between parts, overall joint stiffness, and local stress at the attachment points.

    Analysis type: Static + MPC contacts

  • Slender members under compression

    Critical load factor for a column, strut, panel or thin-walled housing.

    Analysis type: Buckling

  • Vibration-sensitive equipment

    Natural frequencies and mode shapes, checked against the excitation from a motor, pump or transport case.

    Analysis type: Modal

  • Heat-generating enclosures and heatsinks

    Steady-state temperature field, hot spots, and where the heat path is actually restricted.

    Analysis type: Heat conduction

  • Parts with mixed materials or heating

    Thermal expansion mismatch and the stresses a temperature field induces in the structure.

    Analysis type: Thermoelasticity

  • Thickness, rib layout and material trade-offs

    The same geometry in several variants, compared side by side on one scale.

    Analysis type: Linear static, repeated

  • A second opinion on an existing result

    You already have a solution from another package and want an independent run on the same geometry and loads.

    Analysis type: Any of the above

Capabilities

What the solver does today

This is the complete list as of today. If something isn't here, it isn't available yet — and we'd rather you know that now than find out mid-pilot.

WebCAE analysis capabilities, current status and notes
STEP importAvailableAP203, AP214, AP242. Multi-body assemblies supported.
Automatic tetrahedral meshingAvailableSecond-order tetrahedra with curvature-driven sizing.
Local mesh refinementAvailableManual refinement at fillets, holes and contact zones.
Linear static analysisAvailableDisplacements, strains, von Mises and principal stress, safety factor.
BucklingAvailableLinear eigenvalue buckling with critical load factor.
Modal analysisAvailableNatural frequencies and mode shapes, free or constrained.
Heat conductionAvailableSteady-state conduction: temperature field and heat flux.
ThermoelasticityAvailableStress from an imported or computed temperature field.
MPC contactsAvailableBonded, multi-point constraint coupling between parts.
Interactive postprocessorAvailableSection cuts, probes, deformed shape, isosurfaces, animated mode shapes.
AI workflow via MCPAvailableA Model Context Protocol server — drive setup and post-processing from an AI agent.
Workflow

How the pilot works

Seven steps. Two of them need your time — around an hour in total. Everything runs over email, so you can review at your own pace.

  1. You send one or two STEP assemblies

    Export STEP and tell us the loads, supports, materials and what you'd like to know. If you need an NDA in place first, just say so.

    Owner: You: ~30 min

    Time: Day 0

  2. We prepare the model

    Defeaturing, geometry repair, material assignment, load cases and boundary conditions — with a note back to you on any assumption we had to make.

    Owner: Us

    Time: 1–3 days

  3. We build the mesh

    Automatic tetrahedral mesh, then local refinement around the stress concentrators, plus a sensitivity check on the critical region.

    Owner: Us

    Time: 1 day

  4. We run the analyses you asked for

    Whichever of the analyses above fit your problem, with the usual sanity checks: reaction-force balance, energy, unit consistency.

    Owner: Us

    Time: 1–2 days

  5. We go through the results with you

    You get a written walkthrough and a link to the live solved model, so your engineers can section it, probe it and check it themselves. Questions go back and forth by email, at whatever pace suits you. If your team would rather talk it through on a call, we're glad to — it's entirely optional.

    Owner: You: ~30 min

    Time: Async

  6. We share what we'd suggest

    Where the margin looks thin, what's driving it, what we'd change, and what a follow-up analysis would be worth running.

    Owner: Us

    Time: Included

  7. Your engineers keep going in WebCAE — if it's useful

    We hand over the live model, help your team get set up, and stay available. Entirely your call.

    Owner: Optional

    Time: Ongoing

If step 7 turns out to be a no, that's fine — you still keep a report you can use. We'd rather you have a clear answer than a long trial.

Deliverables

What you receive

Everything below is yours to keep, whether or not you carry on with WebCAE.

  1. A PDF analysis report

    Geometry as analysed, material data, load cases, boundary conditions, mesh statistics, solver settings, results, and a plain list of the assumptions we made. Written so someone who wasn't involved can still follow it.

  2. The live WebCAE model

    Not just a screenshot set — the actual solved model in the interactive postprocessor. Section cuts, probes, deformed shape, animated modes. Shareable by link inside your team.

  3. Result fields you can reuse

    Stress, displacement, temperature and reaction data exported for your own post-processing or for comparison against your existing solver.

  4. A mesh sensitivity note

    What changed between mesh levels in the critical region, so you can see how much of the number is physics and how much is discretisation.

  5. Engineering recommendations

    Concrete ones: which feature drives the peak, what a thickness or fillet change does, and where a follow-up run would earn its keep.

  6. A written scope statement

    What we didn't cover and why — so the report doesn't get read for more than it says.

Example

Example case — bracket assembly under service load

Internal example on public geometry, shown to illustrate the format of a pilot deliverable.

Input

A three-part welded steel bracket, supplied as a STEP AP214 assembly. The questions: peak stress under service load, safety factor against yield, and the first three natural frequencies.

Image placeholder

CAD geometry as received. Neutral grey shading, isometric view, no annotations, dark background. 1600×1000.

Fig. 1 — Geometry as supplied (STEP AP214).

Preparation and mesh

Small fasteners and cosmetic features were removed. Parts were coupled with MPC contacts at the weld interfaces. Automatic tetrahedral mesh, with local refinement at the two fillets identified as stress concentrators.

Image placeholder

Mesh close-up on the refined fillet, wireframe over surface, size gradient clearly visible. 1600×1000.

Fig. 2 — Local refinement at the lower fillet.

Results

Peak von Mises stress at the lower fillet, with the resulting safety factor against yield; maximum displacement at the free end; and the first natural frequency compared against the drive unit's excitation frequency.

Image placeholder

Von Mises field on deformed shape. Blue→yellow→red scale matching the site's stress tokens, legend with units visible in frame. 1600×1000.

Fig. 3 — Von Mises stress on the deformed shape.

Mesh sensitivity

Refining the critical region changed the peak stress by only a small margin between the last two mesh levels, which is what tells you the result is converged enough for the decision being made.

Image placeholder

Small line chart, peak stress vs element count, plotted on the site's dark surface colour. 1200×700.

Fig. 4 — Peak stress vs element count.

What we'd suggest

Increasing the lower fillet radius reduces the peak stress at no mass penalty. The upper rib carries very little load and looks like a candidate for removal — worth a follow-up run before committing.

Solver accuracy is documented separately: WebCAE is checked against NAFEMS and MacNeal–Harder benchmarks, with the error against each reference solution published case by case.

Browse the verification suite
FAQ

Questions

How much does the pilot cost?

Nothing, and there's no obligation afterwards. We run it because evaluating a solver on a public demo part tells you very little — we'd rather be judged on your geometry.

How long does it take?

Usually 5–10 working days from receiving the STEP file to sending you the results, depending on the size of the assembly and how many load cases you'd like. On your side it's about an hour in total: a short scope exchange at the start, and reading through the results at the end.

Do we have to get on a call?

No. The pilot is designed to run entirely in writing — you get a written walkthrough plus the live model by link, and we answer questions by email whenever it suits you. Many teams prefer it that way, since it leaves everyone a written record. If you'd like a call, we're happy to arrange one.

Do you sign an NDA?

Yes. Send us yours and we'll sign it before any file is transferred, or we can provide one. Pilot models aren't published, used as case studies, or shown to anyone outside the engineers working on your model without your written permission. If you'd like the files deleted after the pilot, just ask and we'll confirm in writing once it's done.

What CAD formats are supported?

STEP — AP203, AP214 or AP242 — which every major CAD system exports. If you can only supply a native format, tell us which one and we'll confirm whether we can handle the conversion before you send anything over.

Can we use our own models?

That's really the whole point — we don't run the pilot on sample parts. Bring the assembly you actually care about. Ideally one you already have a result or a physical test for, so you have something to compare against.

Do we need to install anything?

No. WebCAE runs in the browser — current versions of Chrome, Edge, Firefox or Safari. Nothing to install, no admin rights, no IT ticket. Larger assemblies are run on our desktop build during the pilot, and you still review the results in the browser.

Who actually runs the analysis?

Our engineers. We write up every modelling decision — defeaturing, contacts, boundary conditions, mesh — so your team can look at the setup and tell us whether they'd have done it the same way. If you disagree with an assumption, we'll re-run it.

What happens to our model and results afterwards?

They stay yours. You keep the report, the exported fields and the live model. If you decide not to continue, you owe us nothing, and we'll delete the data on request.

How do we know the solver is right?

You can check for yourself. Our verification suite publishes results against NAFEMS and MacNeal–Harder benchmarks with the error against each reference solution. Beyond that, the best test is the one you control: run the pilot model through your existing package, or against a physical test, and compare.

What if our model is outside what you can do?

We'll tell you when we first look at it, before you've invested any time. If part of your problem fits — a static check, a modal check — we'll run that part and be clear about what wasn't covered.

Apply

Send us a model

Pick one assembly you're unsure about. Tell us what loads it sees and what you'd like to know, and we'll come back within one working day on whether it's a good fit — and say so plainly if it isn't.

Analyses you're interested in

Optional — pick any that apply, or leave it blank if you're not sure yet.

Please don't attach CAD to this form. Once we've confirmed the scope — and signed an NDA if you need one — we'll send you a secure upload link.

Prefer plain email? info@webcae.com

We keep the intake limited so turnaround stays honest. If the queue is full, we'll tell you the actual wait rather than put you on a list.

We use essential cookies to run this site, plus analytics cookies if you agree. Privacy Policy