If you missed the previous articles, make sure you start there and catch up with the details:
No matter how sophisticated the CAD software, how big the monitor, nor how cerebral the operator, nothing beats a full size mock up you can actually sit in when it comes to fitting in the driver and evaluating the cockpit templates and ergonomics.
A few hours with some plastic conduit and a hot glue gun, some thin sheets of MDF and tape and you will be able to solve all your problems. In addition, if made neatly and then painted you’ll be able to impress some sponsors and University staff. You can probably fool your parents too!
Takeaway 12: make a mock up. It’s cheap and safe!
Stress, CAD, FEA and all those magical optimisation programmes!
It is ultra-rare to see failures due to a one-off load condition: most will be fatigue failures. All stress analysis requires a few fundamental facts, i.e. the load cases being considered, the material properties and the component form.
As a first-year team you will not know the load cases in exact detail, in some cases you might be lucky to be within a factor of 5! You probably won’t know the relevant material properties (i.e. as welded, as machined, as heat treated, as pressed together) to perfection either. And it is quite likely that you won’t have fully modelled local stress raisers such as circlip groove undercuts, surface finish etc.
Consequently, it will not be possible to convince us Judges that “the front upright was the subject of topology optimisation and additive manufacturing was therefore the most suitable manufacturing method using sintered titanium powder etc.”
You don’t need to know exactly the location of all the suspension points to make a start on the chassis structure basics. It does help if you know roughly the angle of the main forces “coming in” from the wheel assemblies so you can apply reasonably representative loads.
As a first-year team all your load estimates will be guesses, possibly well-meaning and educated, but guesses nonetheless. The outputs of any analysis (whether classical methods by hand, using Excel or the best software in the world) will only be as relevant as the accuracy of those guesses.
A simple model of a tube chassis can be very helpful in seeing the main effects of the placement of tubes or shear panels. The change in stiffness, or perhaps more usefully the change in specific stiffness (i.e. stiffness per unit mass) will be very educational: just don’t believe the absolute numbers too much.
To a point, similar practical observations can be made from building a balsa wood model, or a plastic pipe mock up. Twisting such a “structure” can be worth hours spent staring at the screen.
Be very careful with the restraints used, either in a CAD model or in real life as it is remarkably easy to over constrain the structure giving results perhaps 10 times stiffer than they should be… Remember when twisting a box that the faces will warp and lozenge and this should be freely permitted.
Significant local point loads will need careful consideration separate from the global stiffness/mass deliberations. The main ones will be seat belt mounts, engine/differential mounts, lower wishbone mounts and upper spring or rocker pivot mounts. Be careful to understand how the wheel loads are resolved by the time they reach the structure through the angled damper unit or pushrods and rockers noting that the rocker type arrangement can mean significantly higher local forces. The other main focus should be on a stiff and strong mounting for the pedal box; usually complicated by the need to provide a lot of fore/aft adjustment to accommodate the range of driver sizes (still easier than moving the steering wheel!).
On the subject of the pedal box I was disappointed last year to see more than one team shy away from designing their own pedal box. As a first design exercise the pedal box is a great thing to start with: the loads are pretty much two dimensional and can be analysed using simple classical mechanics methods and we even provide you an idea of a load case in the Rules (brake pedal).
I would personally think about fabricating all the pedals and pedal box from thin sheet steel (you wouldn’t need anything more than a holesaw and jigsaw) and I would be looking at car seat base sliding rails as a base mechanism for adjustment (seats with integral belts also place high bending loads into the rails) with maybe a gas strut/spring for pushing the pedal box assembly back to the driver and a cable operating the latch.
Takeaway 13: focus on incorporating good fundamental engineering principles such as straight load paths, removal of bending loads, placing point loads at naturally stiff “nodes”, good weld area, welds in shear, etc.
The Design Freeze
One thing I haven’t mentioned yet is timescales. Time is one resource no-one has enough of. As a first-year team you will be especially tempted to keep on refining as you learn more and become increasingly confident in your progress.
But you need to be mega-disciplined and actually accept that you need to stop theorising and actually start making. Professionals call that the “design freeze” and it’s the date after which no changes are made unless it’s an emergency.
You need to set that design freeze date much earlier than you ever imagine because if you thought learning how to draw, package, design and analyse was hard, then as the song says “you ain’t seen nothing yet”.
Takeaway 14: perfection tomorrow always loses to “good enough” today.
To be a star first-year team all you need to do is get the car finished early so you have time to test, develop/fix and make it reliable. The only way to do this is to get it built early. The only way to have it built early is to start making it early. To make it requires knowing what you are making: see above!
If you have never really made anything, stripped or rebuilt a mechanical device etc. before, then you need a long time for manufacture and assembly. I genuinely suggest you finish your design and preferably drawing processes, for the main areas, before Christmas.
You really don’t need a lot of sophisticated equipment to make a simple but well-designed tube frame FS car. The essentials are listed, in priority order, below:
- Space, preferably dry, with 13A electricity supply. 6m x 3m is probably the minimum.
- Simple bench with decent vice (bench can be home made from wood)
- Tape measure
- Long spirit level (at least 1.5m long)
- Homemade flat surface, piece of 40mm kitchen counter top is ideal, large enough for your main chassis structure
- Good scriber
- Right angle square (larger is better, you can make one from flat steel strip as a 3, 4 5 triangle)
- Metal hacksaw (a jigsaw with metal blades is handy but not essential)
- Flat and half round metal files
- Drill (pillar is handy, good cordless will suffice)
- Drill bits
- Marker pen
- G clamps or similar
- Basic hand tools, hammer, spanners, screwdrivers and socket set
- MIG welder (you can hire one, small enough to work off 13A household supply socket will be fine)
can get away without an angle grinder (cheap and handy though) and a bench folder, air tools and a milling machine but I admit access to a lathe is really helpful.
The metal thicknesses involved are easily cut by hand, including the tubes and shaped with a hand file. You don’t need to get tubes extracted from the CAD model and laser cut. You can make one set of the various brackets for suspension mounts by hand.
If you carefully consider how you might make things at the design stage then you can simplify the jigging. This will involve some compromise. For example, if all your suspension pivot axes are horizontal (parallel to the main chassis base/datum) and you can have them on vertical bulkheads then it becomes a simple matter to make each such bulkhead flat on your table.
Print each bulkhead drawing, complete with suspension brackets full size and glue to your table top. Drill (squarely) through at the suspension pivot centres. Using suitable spacers, you can then build up from the table surface, i.e. bracket, spacer representing the required gap, bracket. Shim the tube heights as required (usually the bracket thickness) and you’re done.
You can then stand up all your bulkheads (like slices of toast in a toast rack) the correct longitudinal distance apart, centred and square to the datum structure centreline and then join then with the requisite side member tubes.
It’s not as elegant or as light or as accurate as making the structure first and adding the suspension brackets after but it will be perfectly adequate. More to the point, it’s really cheap, tooling free and quick, literally a couple of man day’s work (genuinely) to make your main structure: assuming you thought it through up front.
You can use the same sort of approach to locate angled rocker pivots by making a simple right-angle triangle stand (maybe from MDF sheet) with the rocker pivot bolt located at the apex at the appropriate angle and height and slide this up to the side of your structure.
Remember you’re only making one (except perhaps for wishbones where if you have been thoughtful, they won’t be handed) so the ultimate in jig sophistication and durability is not needed.
Takeaway 15: getting things in the correct location with reasonable accuracy is more important than achieving the wrong location with absolute accuracy!
If you have a little bit of cash, and a bit more time planned into your schedule, then you could get a simple sheet steel baseplate water or laser cut to make things a little more accurate and less error prone. Thin hot rolled mild steel sheet around 1 to 1.6mm thick should be quite cheap, you can include small holes to act as locations and jigging references from which you can lay out parallel and mutually orthogonal reference lines from which to work.
This will save measuring with a tape, setting out lines square or parallel etc. With some forethought you can include locational references to make jigging of suspension points etc. and in any dead space (e.g. driver area) you could layout wishbone jig locations?
This steel surface will also prevent burning of the table surface and act as an earth for welding. If laser cut then some firms would be able to etch lines directly onto the surface for you…
If you are having to have parts made externally, even by sponsors or in house but by other people, you need to factor in the lead time. It’s always going to be a minimum of 8-10 weeks unless you are paying top dollar (regular cash paying customers will always be ahead of you). If you are bound to use your University’s procurement process, I can bet this will take at least 4 weeks to negotiate, per supplier.
As always, the simpler you can design the part, the easier it is to draw or specify and the cheaper/quicker it will be to make or buy it. It follows therefore that having some understanding of the likely manufacturing method is a distinct advantage! It further follows that you need to also envisage the likely material required. And again, as beginners, this will of necessity be a learning curve and the steeper part of that curve to boot!
Takeaway 16: wherever possible, the “designer” should be responsible for manufacture/procurement/assembly of “their” parts. And any warranty claims!