On Fabrication

If you are a racer, using a production car as raw material, at some point you are going to identify the need for some go-fast part that the OEM doesn't make. For autocrossers and road racers, this is normally a suspension part. If you have Done the math and have figured out that you need to change the spring rate, or add a thicker roll bar, or change shock valving - it is unlikely that these parts can be found in OEM inventory. Sometimes you get lucky, but for the most part, you are going to need some sort of custom solution.

With that being the case, you have four basic choices:

  1. Buy Aftermarket Off The Shelf - Fire up Google, prowl the web forums, read some magazines, and find & buy an already-existing aftermarket part that fills the need. This is the simplest solution, as all it takes is a credit card, but it is very rarely the right one. The core problem is this: real racers - especially racers who turn corners - are a very, very small market. Spread this market over all the different production cars out there, and the odds are incredibly slim that someone out there has done the same math as you, has come to the same conclusion as to need, has taken the trouble to properly design, manufacture, test, and revise the part in question, and is now offering it for sale.

    Usually, an Off The Shelf (OTS) aftermarket part exists because it is in some way similar to a part that company is already making, and they are trying to expand their potential market while only making minor investments in new tooling to accomodate whatever fitment interface differences exist between your part and the part they orginally bought the tooling for. Usually, the standard for production is "it fits" and there is next to no R&D put into the part beyond that basic fitment test.

    In some cases, that really isn't all that bad - especially with engine parts. A bigger exhaust is probably going to make a little more power and doesn't need a whole lot of extra R&D behind it. For things like sway bars, an aftermarket bar is going to be stiffer than stock. You won't have a whole lot of choice on diameter - so your setup will have to work around whatever you can actually get - and you cannot trust that "stiffer is better" or that a front/rear set were actually designed & tested and so work as a set. You'll still have to do your homework. But some OTS parts are workable if you pay attention to what you are getting.

    Others, though, are absolutely fraught with peril. There is a lot of very pretty junk out there.

    This is junk. Pretty is debateable.

    You are more likely to have succcess if you are using a car for which there exists a substancial market base. In realistic terms, that means Corvettes and Miatas. You can get Corvette fitment Penske shocks that are absolutely 100% the real deal (if you buy them from Penske, and not some Super Tuna). That is because there is a large enough market base (and Corvette guys will spend money) to justify real engineers doing real R&D to produce real parts. There are also Real Race Teams who are being funded by outside agencies to do the business for realsies and who may sell real parts to the general public as a way to add a revenue stream, or clear out a warehouse (Hello Pratt & Miller!). There are still plenty of fly-by-night, snake oil vendors out there, but with a car like Corvette or Miata, you can find the real stuff because the economies of the platform mean guys who make real stuff can stay in business and there are enough real racers feeding R&D into the vendors.

    You might also get lucky and discover a vendor who is working with a racer who is actively working on building real parts. Usually, another racer has decided to turd polish and has convinced some vendor that this market exists. You will gain access to the fruits of their R&D - for a while. These sources invariably go dark, because no matter how convincing the racer is and no matter how good the collaberation between them is, the market size for racing 2nd Gen Humpmobiles just isn't there to support this for long. Eventually market realities catch up and the vendor bails.

    The other risk is that you are at the mercy of the engineering skills of Turd Polisher and his supporting vendor. Some are very good; others, not so much. You must be very very suspicious of any OTS part and test the damn thing before you bet the farm on it. You must be prepared to discover than any Super Go Fast bit - no matter how strong its pedigree - may be junk, and the money you spent on it wasted.

    Somtimes, it's not that part is junk per sae, but you (as a racer going around corners) may be using it (and stressing it) in ways not envisioned by the manufacturer. I blew up six transmissions in the Talon. Each time, we tried a different mitigation strategy to try and keep it together and each time it didn't work. I finally did an analysis to figure out how many runs I was getting before boom, and pulled one apart just before we got to boom. That revealed the culprit; it was the "indestructable" Quaife centre diff. Quaife offers a lifetime warranty, so it became team SOP to pull the Quaife out every 65 runs and send it back for a new one - which Quaife honoured every time. That got to be very expensive for Quaife, but the mechanical envelope within the trans did not allow enough room to make the points of failure stronger so there was no way for Quaife to solve the problem. The performance benefits of the part were enough to justify using it - but twice a season, out it came. Junk? No, not at all - but there were limitations with it that needed to be understood.

    Quaife no longer offers a lifetime warrantee on that specific part number....

    Which brings us to:

  2. Build It And They Will Come - Convince a vendor that a market exists for your go-fast part, and get them to build it to your specifications. Usually, you get the part for no financial outlay (the vendor eats that) but you have testing and promotional responsibilities. I did a lot of this with the Talon. I had a lot of success (and a few horrible failures) going this route. The success of my sponsors was somewhat mixed. Nobody got rich based on parts I had a hand in developing. Some of my R&D played a part in the success of other parts, mostly because the vendor learned things from my feedback or on how I broke stuff. For some, the only tangible payoff they saw was ego-boo in seeing their name on a race car and hearing my (very real) gratitude for their support. It can be fun to help a racer achieve success (if the racer holds up his end of the deal).

    One vendor did a ton of R&D on designing a turbo for us. Our most successful season generated zero sales of this part. Then we went to a drag strip, put down a suprising number, and they flew off shelves for a while....

    If you can get a vendor to sign up for this, more power to you. Be very, very careful about who you hitch your star to, because there are some seriously shady characters out there, and don't be afraid to fire a sponsor if things aren't working out.

    The success of this route is very time & space dependant. You have to be in the right car, at the right time, with the right vendor, to be able to make this work. I caught the rising tide of the whole import car craze and happened to be right in the thick of a specific community that made my approach work. It was doable in 1998. I couldn't duplicate it in 2016, were I to try - not with that specific car anyway. The mechanisms I exploited for promotion simply no longer exist. You absolutely cannot bank on this approach working and a skeptical vendor is a smart vendor.

  3. Pay Someone to Build It - you do the design and engineering, somebody else builds it. In North America, especially in Rust Belt cities, there exists a substancial amount of manufacturing capability. Welding shops, sheet metal shops, machine shops, metal foundaries, anodizers & polishers... ... all these small-to-medium size job shops who produce little of their own designs, but instead act as sub-contractors for other firms who leverage their capablities to produce end products. These shops absolutely can be leveraged to produce whatever parts you need.

    There are two major obstacles to overcome:

    1. The first is that you need to be able to generate accurate and legible mechanical drawings. The shop will have absolutely no idea what a "camber plate for a 2G Humpmobile" is, or does. They will not be able to spot errors in your design from a "part works as intended" point of view. They will be able to spot "I can't make that" errors, or areas where the drawing is confusing or unclear - but once that is sorted out, the drawing is the contract, and they will deliver exactly what is on the drawing. If that doesn't fit or doesn't work, that's your problem as the engineer.

      An example mechanical drawing
    2. The second is that you will probably pay through the nose, because your parts don't offer the economies of scale that they are used to working with. The normal process of quoting a job bills for quote time, drawing analysis time, fixture production, and machine set-up time as fixed costs. On top of this, add material costs and machine operating costs (machine time, operator time, and consumables) per part, For reasonable part complexity and material choice, the cost per part is normally pretty cheap; it is the fixed costs that hurt. So this means that it is normally almost the same price to run a dozen parts as it is to run 50.

      The other aspect that once your job has been worked out, the shop will keep notes on how they did it so subsequent part runs have less analysis. The best case for one of these jobber shops is a phone call "I need another 200 of part number ABC123". The shop pulls the fixture for ABC123 out of the closet and goes to town on the job.

      You are not in that business. You won't want regular production runs of 200 units - you want 2 units (gotta have a spare), once.

      Most job shops don't want that business, as they make their money on volume. So small-volume, short runs tend to be quoted with "let's see if he is serious" pricing - which isn't cheap.

      Sometimes, you get lucky. Sometimes, if you can be flexible on delivery you can be mixed in with another job (especially batch jobs like anodizing and plating), or maybe the shop will give your job to the new apprentice using that cranky old Bridgeport in the back corner or whatever.

      The one exception these days is the explosion of consumer/hobby CNC machines has opened up a market of Internet machine shops. There are now places where you can submit CAD drawings and solid models, and various operators will bid on making the part. This can be really cheap, as often the payoff for the "shop" is being able to build something, not the actual money paid for the job. Quality and (especially) delivery time can be all over the map though, and they usually can't deliver on finishing like anodizing.

      An important consideration when you are designing parts for someone else to make is you must consider not just how the finished part works, but how the shop will actually build the part. Anything you can do in the design to keep the part simple, so it can be made with the fewest setups on the most basic machines will increase the odds of the shop taking on the job and reduce the price. If your part must be made in the 6-axis machining centre, you will be competing for machine time with high-volume production jobs and you will pay for the privledge! Make friends with your subcontractors!

    I had some success with this technique, but I was lucky enough to know how to produce really good drawings, and I was willing to horsetrade and work with the shop to meet halfway. Mostly, I used this for capabilities I didn't have at the time, like welding, anodizing, and plating. It was so good to live in Windsor / Detroit, where I had so much choice in suppliers! Rural New Brunswick, not so much....

  4. Build It Yourself - the only way to do it right is do it yourself. The primary attraction of this method is that you are master of your own domain. The primary disadvantage is that you are master of your own domain. Success means developing not just the engineering skills to design your own parts, but the fabrication skills to be able to convert your ideas into reality.

    This takes time and money. You will need a lot of tools, and you need space to use them.

    For most of Far North Racing's active operation time, we lived in a 2-bedroom apartment on the 9th floor in downtown Windsor. That second bedroom was the fabrication shop. I had a drill press, a belt sander, an aluminium-cutting chop saw, a Smithy mill-drill, and a small air compressor in there. I used that room to revalve shocks, I built intake piping in there, I built an intercooler in there, I fabricated dozens of brackets, spacers, and adaptors....

    For bigger jobs, I had access to a friend whose day job was a tool & die maker who had a huge machine shop on his farm. If I had a lathe or mill job that wouldn't fit on my Smithy, he'd let me use his Bridgeport or South Bend.

    These days, I have my own rural shop, with welders, a lift, a blasting cabinet... you get the idea. I have burned a lot of money on tools, and it isn't likely to stop any time soon. But every tool adds capability that I didn't have before, and most jobs these days I can do without having to buy a tool to make it work. It is very, very nice to be able to go from "good idea" to "part in hand" in a few hours of work.

    Part of the trick is to avoid re-inventing the wheel whenever possible. Most purpose-built race cars (i.e. cars that were built from scratch, rather than modified production cars) use commodity parts whenever possible. Springs, for example, come in a couple of set form factors in a wide variety of different lengths and rates, and the companies that build them can make use of economies of scale to push quality up and cost down. You will never be able to fabricate a suspension spring at the level of quality and low cost that Hypercoil can - so don't try. Instead, your job is to figure out how to adapt commodity race car parts to interface with your production-based car.

    Adapting OTS AP Brake parts to a C5 Corvette

    The majority of my racing engineering ouvre was identifying real race car parts and designing adaptors that fit them to the car I was working on. Not only is this strategy cheaper in the long run, it can be very helpful to have commodity parts on the car for when stuff breaks or when you want to sell off unused stock. You aren't running down to Averill Racing on a Saturday to pick up some bespoke part, but a 2.25" x 7" x 500lb Hypercoil or an AP Racing master cylinder you might be able to get.

Supplier catalogs are an engineer's best friend! Every year I'd load up on material from PRI and that info was invaluable.

Of all the fabrication strategies, "Do it Yourself" was my favourite, but I made use of all four. Not all of us will have access to a killer shop (or have a wife so understanding as to let you build one in the second bedroom of your two room apartment!). Notwithstanding, the more hands-on you can get yourself, the better you will understand your car, the better your parts will get, you will have a better feel for differentiating between "wheat" and "chaff" when it comes to OTS parts, and you will be able to talk to subcontractors more intelligently (and so improve your chances of getting what you want at a decent price).

No matter what strategy you take though, you must always always always test every single part as conclusively as you can. Any vendor who wants to keep something secret from you (as a lot of Super Tuna engine tuners and shock revalvers seem to want to do) is not "protecting his intellectual property"; he is hiding his shoddy work. Never accept a pig in a poke!