Overview
Objective Take the Concept Class chassis from concept design through build and in-house track testing, meeting stiffness, packaging and cost targets inside a fixed competition deadline.
Concept Class (Class 2) is judged on the design and the reasoning behind it rather than on a finished car's dynamic performance, which puts the weight on being able to show why the chassis is the way it is. My work covered the frame and body: three full design iterations in SOLIDWORKS, regulation compliance checking against the FSUK technical regulations ahead of the Engineering Design and Cost & Manufacturing submissions, and support through build and in-house track testing.
The most instructive part was not the CAD. During in-house testing the car showed an instability under acceleration. Rather than treat it as a suspension setup problem, we traced it to compliance in the rear engine mounting — the mount was deflecting enough under drive torque to change what the rear of the car was doing. Revising the mounting for local stiffness resolved it, confirmed by driver feedback across repeated runs, and the revision was worked through with manufacturing so it stayed buildable inside the deadline.
Design brief, requirements and constraints
A Concept Class chassis has to satisfy three things that pull against each other: it has to be stiff enough that the suspension geometry behaves as designed, it has to package the driver, powertrain and systems inside the regulation envelope, and it has to be affordable and manufacturable by a student team inside one season.
Everything else on this page follows from those three constraints and the fact that the competition deadline does not move.
| ID | Requirement | Why | How verified |
|---|---|---|---|
| R1 | Frame and body compliant with the FSUK technical regulations for the class. | Non-compliance is not a penalty, it is exclusion from the event. | Rule-by-rule check against the regulations before each submission. |
| R2 | Torsional stiffness sufficient that suspension geometry behaves as designed. | A frame that twists significantly under cornering load changes the effective geometry, so the setup work is chasing a moving target. | TO CONFIRM — state the target value and how it was checked. |
| R3 | Package driver, powertrain and systems inside the regulation envelope. | Packaging conflicts found after manufacture are expensive in a fixed-deadline season. | Assembly-level interference and clearance checks in SOLIDWORKS. |
| R4 | Buildable with the team's available manufacturing capability and within the cost target. | A design the team cannot build or afford does not score in Cost & Manufacturing. | Review with manufacturing before each iteration was frozen; Cost & Manufacturing submission. |
Constraints
- Fixed competition deadline — the date does not move for a redesign.
- Concept Class (Class 2) regulations and the FSUK technical regulations.
- Student-team manufacturing capability and cost target.
Out of scope
- Suspension kinematics design.
- Powertrain and battery system design.
Design decisions and rationale
Each entry states what was chosen, what else was on the table, and why the alternative was rejected.
-
Treat the instability as a structural problem, not a setup problem
Decision When the car showed an instability under acceleration during in-house testing, investigate the load path through the rear structure before changing suspension setup.
Alternatives considered
- Change suspension setup — damper settings, geometry, tyre pressures The usual first response, and it would probably have masked some of the symptom. But the instability appeared specifically under acceleration, which points at where drive torque is reacted rather than at how the car is sprung. Setup changes would have consumed test days without addressing the cause.
- Treat it as a driver technique issue It was repeatable across runs, which is the signal that it is the car and not the driver.
Why this one
The symptom correlated with a specific load case rather than with speed or corner phase. Under acceleration, drive torque is reacted through the engine mounts into the rear of the frame. That made the rear mounting the first thing to check, and inspection found enough compliance there to explain what the driver was reporting.
The general point: a symptom that appears under one specific load case is telling you where to look, and it is worth reading that before reaching for the setup sheet.
Trade-off accepted Time spent investigating the structure before trying the cheaper setup changes — which would have been the wrong call had the structure been sound.
Evidence: Repeatable driver reports under acceleration, and inspection of the rear engine mounting.
-
Stiffen the mounting locally rather than redesign the rear frame
Decision Revise the rear engine mounting for local stiffness, leaving the rest of the frame architecture unchanged.
Alternatives considered
- Redesign the rear frame bay around the powertrain Would have addressed the compliance, but the manufacturing time and cost were not available inside the remaining schedule, and it would have invalidated the packaging and cost work already submitted.
- Add a full bracing structure across the rear Heavier and more material for stiffness that was not needed globally. The compliance was local to the mounting, so the fix should be local too.
- Accept it and manage it through setup Leaves a known structural deficiency in the car and makes every subsequent setup change harder to interpret.
Why this one
The measurement of the problem was local, so the smallest change that addressed it was local. That kept the revision inside what manufacturing could produce in the time remaining, and left the packaging, cost and regulation work already completed still valid.
Trade-off accepted A local stiffness increase raises the stiffness discontinuity at the mounting, so load is concentrated more sharply into the surrounding structure. Worth checking in any future iteration.
Evidence: Driver feedback across subsequent runs after the revision.
-
Coordinate the revision with manufacturing before freezing it
Decision Work the revised mounting through with the manufacturing side of the team before committing to it, so that buildability was a design input rather than a discovery.
Alternatives considered
- Finish the design, then hand it over to be built On a fixed deadline, a design that comes back as unbuildable costs the whole loop again — and there was not another loop available.
Why this one
With no schedule slack, manufacturability is a hard constraint, not a preference. Bringing it in early cost a short conversation and removed the risk of a late redesign.
Trade-off accepted Some design freedom given up in exchange for certainty that the part could be made on time.
-
Validate through driver feedback rather than instrumented measurement
Decision Confirm the fix using structured driver feedback across repeated runs under the same conditions that produced the original symptom.
Alternatives considered
- Strain gauges on the mounting, or accelerometer and IMU logging Would have given a quantified before-and-after rather than a qualitative one. Not available within the equipment and time budget for the test sessions.
Why this one
Driver feedback is subjective, but it is not worthless: the original symptom was reported repeatably by the driver under a specific and reproducible load case, so the same driver reporting it gone under the same conditions across several runs is meaningful evidence.
It is qualitative evidence, and it is stated as qualitative rather than dressed up as a measurement.
Trade-off accepted No numerical stiffness improvement can be quoted for the revision. That is the honest limitation of the test method, and it is the first thing I would change with another season.
Engineering logic and methodology
- Concept and packaging. Frame architecture laid out in SOLIDWORKS around the driver, powertrain and systems package, inside the regulation envelope.
- Regulation compliance. Frame and body checked rule-by-rule against the FSUK technical regulations, ahead of the Engineering Design and the Cost & Manufacturing submissions.
- Iteration. Three full design iterations, each one reviewed against the stiffness, packaging and cost targets, and each one checked with manufacturing for buildability before being frozen.
- Build. Manufacture and assembly, with design support for the issues that only appear when parts meet each other.
- In-house track testing. Running the car and collecting structured driver feedback against specific load cases.
- Diagnose and revise. Instability under acceleration traced to rear engine mounting compliance; mounting revised for local stiffness; revision agreed with manufacturing.
- Re-test. Same driver, same conditions, repeated runs, to confirm the symptom was gone rather than moved.
Calculations
Torsional stiffness — the quantity a chassis target is written in
Chassis stiffness targets are quoted as torsional stiffness, and it is worth being explicit about what is being measured, because a number quoted without the test arrangement is not comparable to anyone else's.
Assumptions
- Linear elastic response over the load range applied — check the load-deflection plot is straight before quoting a single number.
- Frame constrained at the rear suspension pick-up points, loaded at the front pick-ups.
- Suspension and tyres removed or rigidly replaced, so the number describes the frame and not the vehicle.
| Symbol | Meaning | Value | Unit | Source |
|---|---|---|---|---|
| T | Applied torque about the longitudinal axis | \mathrm{N\,m} | ||
| F | Force applied at each end of the loading beam | \mathrm{N} | ||
| b | Track width of the loading points | \mathrm{m} | ||
| \delta | Vertical deflection measured at a loading point | \mathrm{m} | ||
| \theta | Twist angle of the frame | \mathrm{deg} | ||
| K_t | Torsional stiffness | \mathrm{N\,m/deg} |
-
A twist test applies equal and opposite vertical forces at the two front pick-up points, separated by the track width, giving a pure torque about the longitudinal axis.
T = F \times b -
The twist angle comes from the vertical deflections measured at the two loading points, using the small-angle relation.
\theta = \arctan\!\left(\frac{\delta_{1} + \delta_{2}}{b}\right) \times \frac{180}{\pi} -
Combining the two gives stiffness directly from the measured quantities.
K_t = \frac{F\,b}{\theta}TO CONFIRM — substitute the applied load, track width and measured deflections from your test.
Result TO CONFIRM — state the measured or predicted torsional stiffness, and the target it was assessed against.
Quote the test arrangement alongside the number. The same frame measured with different constraint points produces a materially different figure, which is why bare stiffness numbers between teams are rarely comparable.
Local stiffness at the engine mounting
The rear mounting problem was a local stiffness problem, so it is worth writing down the quantity that was actually deficient. Torsional stiffness of the whole frame does not capture it.
Assumptions
- Deflection measured at the mounting under an applied load representative of the drive-torque reaction.
- Response linear over the load range applied.
| Symbol | Meaning | Value | Unit | Source |
|---|---|---|---|---|
| F_m | Load applied at the mounting | \mathrm{N} | ||
| \delta_m | Deflection at the mounting under that load | \mathrm{m} | ||
| k_m | Local stiffness at the mounting | \mathrm{N/m} | ||
| M_d | Drive torque reacted through the mountings | \mathrm{N\,m} | ||
| r | Effective lever arm from the crank axis to the mounting | \mathrm{m} |
-
The load at the mounting under acceleration comes from reacting drive torque across the mounting positions.
F_m \approx \frac{M_d}{r} -
Compliance is simply the inverse of stiffness — it is the quantity the driver actually felt, because it is deflection per unit load.
C_m = \frac{1}{k_m} = \frac{\delta_m}{F_m} -
The revision aimed to raise k_m, and therefore cut deflection under the same reacted load.
\delta_{m,\text{revised}} = \frac{F_m}{k_{m,\text{revised}}}TO CONFIRM — if you have before-and-after FEA or measured deflections, put them here. If you do not, say so rather than estimating.
Result The revision targeted local stiffness at the mounting rather than global frame stiffness, because that is where the compliance was.
This calculation is written symbolically on purpose. No before-and-after stiffness was measured on the car, so quoting one would be inventing it.
Validation, testing and what was learned
Validation was in-house track testing rather than a rig or instrumented measurement, which sets the limits on what can be claimed.
The original symptom was repeatable: the same driver reported instability under acceleration across multiple runs. That repeatability is what made it a usable test signal. After the mounting revision, the same driver ran the same conditions again across subsequent runs, reporting on the same specific load case rather than on the car in general.
The instability under acceleration did not return across the subsequent runs. The evidence is qualitative — structured driver feedback under a controlled and repeatable load case — and it is worth being clear that this is not the same as a measured stiffness improvement.
What it does establish is that the change addressed the symptom the driver could feel, in the load case where it appeared, repeatably.
What went wrong
- The instability was found during in-house testing, not during design. A compliance check on the rear mounting load path at design stage would have caught it before it was built.
- The test method was qualitative. Without instrumentation there is no before-and-after number for the revision, only the absence of the symptom.
- The fix was made under deadline pressure, which constrained the solution space to what manufacturing could produce in the time remaining rather than to the best available option.
What I would do differently
- A symptom tied to one specific load case is pointing at a load path. Read that before reaching for setup changes.
- Bring manufacturing into a revision before freezing it, not after. On a fixed deadline, buildability is a design constraint like any other.
- Design the instrumentation into the test plan. A strain gauge on the mounting would have turned a qualitative result into a quantified one for very little cost.
- Check local stiffness at hard points that react torque, not just global torsional stiffness. The global number would not have shown this problem.
Next Instrument the rear mounting and repeat the test to quantify the improvement, and add a local compliance check on torque-reacting hard points to the design review checklist for the next iteration.