Time for a little wild eyed venture into the epistemology of upgrading automobile components and comparing various brands and such of approximately stock components. Many times I have heard "gee whiz" reports on items of this type, like:
Remember your science, folks. Measurement must be accurate and objective, before and after changing just one variable, under controlled conditions. Measurements should be averages, not single observations. My main criticism of many anecdotal reports is that the parts in question were usually being changed because the old ones wore out. The person then looks around at various stock and upgrade options, chooses one, and reports on it. Having swapped several good parts for different good parts in the last year (struts that didn't have spacers for another pair; brake discs, calipers and rotors; and a few others) and noticing no difference between them got me to thinking, and I will express the dubious results of this thinking on a system by system basis. Air filters So you replaced your 40,000 mile old OEM filter with a fancy (brand deleted) unit that cost twice as much and promised "up to X%" horsepower gains? Hmmm, well even zero is "up to" any percentage, but we'll minimise that critical thinking approach for a moment. When was the last time someone yanked out a new filter they put in last week and put in an aftermarket unit? This would be the comparison that would make sense. The "new trick filter for filthy old one" swap is bound to be an improvement, especially with the psychological enhancement of the extra money spent, pseudo-science in the product literature, and bragging rights about the "upgrade." Some of these fancy filters, which don't filter as well to get their effect - basically they are more open, which impedes the air flow less, but increases the size of the particles allowed through them - result in a louder engine sound. This is because the air filter also filters sound coming out of the engine through the intake manifold. Less filtration = less noise reduction - and maybe more power - but the louder noise makes the car sound more like a race car... and hence feel faster.  Spark plugs New for old? No comparison. Every set of plugs I have installed in my cars ever has improved the power and smoothness of the engine. Especially the time I ran a set of platinums down to the point where the car would hardly run! Again, and this is my theme in this "essay", what matters is the apples to apples comparison of two different new products in a close time frame, like the same day. Same tank of gas. Same relative humidity. Same amount of wear and misadjustment on every other part. Do this and you will really know what the difference is. Brakes Typical: Someone installs stainless steel, teflon lined brake hoses and is amazed at the improvement in brake pedal feel. When I installed my set I noticed nothing. Of course, I had basically rebuilt my brake system completely (more than once even) just before doing this, which involved flushing the old disgusting brake fluid out and replacing it with fresh. Could it be that the brake line replacement someone does inspired them to perform if not a complete flush (quite likely) at least a good bleeding of the system? Which will improve brake feel on its own... No complaint: reports of comparative brake pad wear and effectiveness by people who swap pads often, due to racing requirements and such. They are usually comparing last weeks new pads with this weeks new pads. No quite scientific but the comparison is much more valid. Suspension components It would be redundant to reiterate my argument about old worn parts vs. new "upgrade" parts. So I won't. But you can figure it out yourself I'm sure. This is the place, however to introduce the "different" = "better" fallacy. Often, a change is perceived as an improvement, whether cosmetic or functional. Why else would we North Americans hanker after cosmetic "euro" parts while in Europe people customise their cars by putting US-only parts on them? Newer model parts on old cars, older model parts on new cars, the change will make you feel good regardless of objective differences. Heck, the Electro-De-Luxe badge on the back of my Vega made it cooler. The reason I am addressing this under suspension bits, is there is some serious engineering involved in this area of a car, or at least there should be. On our Audis, it is one of the best parts of the car. So "different" could be a serious reduction in the validity of the original compromises (handling vs comfort, cornering vs. straight line, etc.) made by the cars designer. I'm not saying a car can't be set up with a different set of suspension components and exhibit an all around improvement, just that not only is it more likely that there will be a trade off significant enough not to be a valid "improvement" (say, 40% reduction in straight line stability for a 25% improvement in cornering ability, as if these things were quantifiable!) but that the "improvement" might not even be one at all. Hey, so long as the driver is happy, that is all that really matters. But lets call things as they they really are. On my car I have a set of lowering springs, a front strut stress bar, really wide tires, and stock struts and shocks which are relatively new. Report: the new struts and shocks were really nice to have. The car, which never even failed the "bounce test", no longer nosed up and down with acceleration and braking. The front stress bar radically changed the feel of the steering. I honestly think this was an improvement, but as someone pointed out, I might have been correcting some body flex due to age and rust between the strut towers and rest of the cars body. How can I ever know for sure? The lowering springs didn't change the feel of the car much, but made it look a lot better. I scrape on speed bumps now, the car is less susceptible to side winds, and did I mention that it looks a lot better? However the combination of this and a set of tires 40 mm wider than stock, have resulted in some rubbing in the rear when carrying a couple of adults over bumpy roads - which luckily I don't do very often. The wide tires pretty much ruined the cars straight line stability. Of course, it's still better than any minivan, but it will wander if I don't pay attention, and will grab and follow depressions in the road surface much more readily. All I have to do is pay attention to avoid this though - is that such a small price to pay? I don't have to fight to keep it straight, just not look over my shoulder or play with the radio much. The car now corners like a wild thing of course. Whereas I used to be able to make the tires screech, and scare myself at the fastest speeds I would take certain curves, now, um, I'm just scared to try to get anywhere near where these tires would start to let go. Too fast for my blood. So is the car "better"? Objectively, I would say certainly not. It's harder to drive (pay attention!) in mundane situations, and may even be tougher to control properly under extreme situations - who knows what I have done to the predictability under emergency manouvering that I always admired in my Audis? I certainly hope I never get a chance to find out, but if I do I will report... Next topic: changing more than one variable. The only legitimate objective test for most of these upgrades is to swap one single new part (or set of parts) for another new set, with before and after data on mileage over a tankful of gas, acceleration under some measureable condition, cornering under some measurable condition, etc. It's tough enough to do this on the street, without the natural tendency to fix something else while a system is apart. So if you swap some new spark plugs in a week after another new set, you have to resist the temptation to touch anything else on the car until your comparison is complete. No cleaning, no moving wires around, no tire inflation, no dry gas, don't even take a shower! Ok, maybe that's little extreme, but you get the point. It certainly can be argued that an un-needed (ie, the old parts are still good) upgrade such as lowering springs and stiffer shocks could be considered "one" part, but who's to tell what their interaction is? Maybe one was a great improvement, the other a detriment, and the combined result a moderate upgrade. But you have no way of telling what the situation really is. It is certainly possible, perhaps even likely, that one is only an improvement in the presence of the other, but the only way to know for sure is to compare good working stock parts to each one on its own and then to the combination of both. Also there is the fact that either of these parts may work better in conjunction with just the right version of the other one. This is a litle more wrenching than most of us want to get into! So we have to rely on the reports of the people who swap parts more often than "necessary", or wear them out really fast - the racers. Who, of course, are not striving for the same combination of qualities that the street driver is. How often do we hear "I want to replace my worn struts and springs with something better, but don't want the ride to be too harsh"? Well, maybe, just maybe, the stock setup with new parts is the best thing that could be on the car! But the new "upgrade" parts will still be better than the old worn out stock parts when the swap is effected. When I swap my camshaft for a turbo engine unit, how will I know what I've done, since I'll be changing the water pump, timing belt, valve stem seals, cam cover gasket, and who knows what else at the same time? Is my old camshaft worn? I probably won't ever know. Will the donor camshaft be up to its original spec? Again, unless I do some really careful measurements I won't even know these facts. If my valve stem seals are leaking air (they're probably leaking oil, where else would it go?) I'll have to retune the car, and it's gonna be better even if the camshaft does nothing! So you really want to know if a modification improves the power of your engine? Before doing it, clean up the car and tune it. Fix the little stuff so it isn't an issue or a variable. Then go out and do some measurements, carefully and safely. For instance, measure 30-50 acceleration in 3rd gear by having a partner with a stopwatch say "go" at 30 and then stop the timer when they see the needle hit 50. Do this at least ten times and average the results. Then install your modification, and go out with the same tank of gas (or refill to the same point from the same gas pump - better yet refill from a container filled at the same time and place as the tank you were running) and the same partner, on the same stretch of road, at the same time of day and in the same weather, and use the same system of measurement and compare the results. Ideally, you would have a machine make the measurements and not look at the data until the entire process is complete, to eliminate observer fluctuations prejudicing the results. I think I made my "point", thanks for bearing with this somewhat ranting incoherent presentation! |
