Direct 'Dear Dave' tech letters to email@example.com.Coleman will share mind-numbing details, earth-shattering revelations, and technical nerdisms in this space each month.
Q Mensa ManI want to point out a common error in your analysis of the Wilwood calipers on Project Sentra SE-R. You stated that: "pushing four 34mm pistons takes about 60 percent more fluid [than the 54mm single-piston NX2000 calipers]." With sliding calipers, the piston area is accounted for twice, since it moves twice as far as in a fixed caliper. Dave must know this already, because he wrote about it correctly in Technobabble, Dec '02: "Third thing to remember: sliding calipers act like they have twice their actual piston area."
The effective piston area of the Wilwood calipers is only about 80 percent of the NX2000 caliper, 3,630 mm2 versus 4,580 mm2. Any pedal feel issues he had must have been related to air in the system. Increasing from a 7/8-inch to 15/16-inch master cylinder isn't enough of a change to account to go from bottoming out to a firm pedal.Cory LongLong Beach, WA
A I think you can get a Mensa membership just for using my own four-year-old quote to prove me wrong. Luckily for me, I wasn't quite as dumb as you think I was.
What Mr. Long is rightfully pointing out is a fundamental difference between fixed and sliding caliper designs. On a four-piston caliper, moving the pads 2mm closer together, for example, requires the inboard and outboard pistons to each move 1mm. With each 34mm piston having an area of 908 mm2, 1mm of movement takes 908 mm3 of fluid. Four pistons means 3,632 mm3 of fluid in that caliper.
On a single-piston sliding caliper, moving the inboard pad also means moving the piston 1mm. The outboard pad, though, is moved by the big claw that reaches around the rotor. That claw is moved when the cylinder moves backward, relative to the piston. Moving the 54mm piston 2mm takes 2,290 mm3 of fluid from piston movement, and another 2,290 mm3 from the cylinder pulling on the claw, for a total of 4,580 mm3.
The math club guys already know 3,632 is not 60 percent more than 4,580. So what was I smoking? A little too much brake geekery, that's what.
Turns out that although it takes twice as much fluid to move a sliding caliper a given distance, the distance it has to move in the first place is only half as far. In a perfect world, with dead straight rotors, and completely rigid hubs and wheel bearings, the distance a brake pad has to travel from its rest position to the rotor surface is determined by the piston seals. When the piston slides toward the rotor, the seal first sticks to the piston and just stretches out about 0.1mm, then the piston starts sliding through the seal. When the pressure behind the piston is released, the seal pulls back by that same 0.1mm. Assuming there's nothing else pushing the piston back, that 0.1mm is how far the piston has to travel next time before the braking starts.
On a four-piston caliper, the pistons have to squeeze 0.2mm, since there are seals pulling back from both sides of the rotor. On a sliding caliper, there is only one seal pulling the piston back 0.1mm, so the free travel of the caliper is shorter.
So, the free play on the 54mm Nissan caliper came from moving the piston and cylinder 0.1mm apart, taking 229 mm3 of fluid. On the 34mm Wilwood, it takes 363 mm3 of fluid to move all four pistons 0.1mm. The grand total, 62 percent more fluid, just like I said.
Only problem is, when I initially made my bold, 60 percent proclamation, I didn't actually know how far the piston seals pulled back. I had assumed it was far more than 0.1mm, and that it was the main contributor to brake free play. Turns out that wavy rotors and flexy mounting will cause most of the piston knockback that occurs between brake applications, so my numbers, in the real world, were complete rubbish.
StiffieQ I have a 1999 Honda Civic hatchback with an Integra Type-R engine and five-bolt spindle/brake set-up. I am currently running Buddy Club racing spec coilovers with 12kg/mm springs in the front and 18kg/mm springs in the rear. I'm running the factory Civic EX anti-roll bar in the front with an A-spec racing 32mm hollow adjustable anti-roll bar in the rear. I plan on running in the CASC (Canadian equivalent to SCCA) Solosprint series which is like the Time Attack format run in the US. My issue is this: a lot of people are telling me to remove both my anti-roll bars, as they will have little or no effect with the spring rates I am running. I have mixed feelings about this, as I've driven my car with and without the anti-roll bars and have felt quite a big difference between them.
I've also noticed that cars like the older Realtime Integra Type-Rs used to run both front and rear bars, even with spring rates higher than mine. Do you suggest I keep the anti-roll bars on, or do you agree with my racemates that I should take them off?John ColmsToronto, Canada
A First, let's calibrate our brains to your spring rates: 12kg/mm is 670lb/in, and 18kg/mm is 1000lb/in (multiply by 55.8 to convert). A Civic's front motion ratio is about 0.7:1, so the wheel rate is 469lb/in., and the rear ratio of 0.75:1 turns your 1000lb/in. springs into 750lb/in. at the wheel.
This is your street car? Ouch.
Your buddies are right to think the suspension is moving so little that the bars won't have a chance to do much. That doesn't mean they aren't doing anything, however.
Though you've already tried and proven that you can feel a difference, you still don't know which set-up is faster. My recommendation is to establish a simple test regimen next time you're at the track. Go out and set a lap time with your current set-up in the first session. In the second session, make a change based on how the car handled the first time out. Too much understeer? Disconnect the front bar. Too much oversteer? Disconnect the rear.
To save time, you can just remove the end link on one side of the bar you're disabling. This will save you the time of actually pulling the bar out. If your lap times improve, you can pull the bar out later to save weight.
In the HoodQ I have a 1997 Accord Coupe with the F22B DOHC non-VTEC engine and a brand new Evo III 16G turbo. It seems that I should try to eliminate under-hood heat as much as possible. Would a vent above the manifold/behind the radiator be a good idea?
I have been going over the question of hood vents for the past few weeks. I've been talking to people on many different websites, I've gone to Mitsubishi dealerships to check out the Evo IX's hood, and even dug back to the article in Sport Compact Car that was written when the SRT-4 came out. I'm trying to decide if I should invest in a vented hood.
To add to the confusion, in your most recent issue, the Miata went on a diet and a hood vent was put in behind the radiator; the new Evo is showing a NACA vent on the back of the hood near the windshield; your Silvia doesn't have anything except the stock Silvia hood; and some magazines show Intergras, 180SXs, etc., with the rear of the hood lifted maybe a quarter of an inch at the hood hinges. Which of these techniques is best? I want to know what you, proven thinker and problem solver of all things, have to say.Mike WildeboerCincinnati, Ohio
A Your goal is a wise one, and the quest for the best solution isn't easy. An exit vent behind the radiator can improve cooling if it's designed so air actually comes out of the vent at speed. Your main concern should probably be under-hood heat from that hot turbo which, presumably, has no heat shielding on it. Here are a few things you should think about:
1: NACA ducts, like the one on the Evo X, are for letting air into the engine bay. The new Evo's turbo is on the back of the engine, and the vent blows air onto the turbo much like the vent on an SRT-4 does. Your turbo is in the front, and there's no room for a NACA duct ahead of your turbo.
2: At speed, the air at the rear of the hood tends to be at relatively high pressure, since it's hitting the windshield and getting forced up over the roof. Lifting the back of the hood to let air out, like the riceboys in those other magazines, is more likely to actually let air in.
3: Getting cooling air across your radiator and intercooler means making sure the air pressure in front of the car is higher than the pressure in the engine bay. If you open the engine bay to the pressurized air at the back of the hood, you'll likely increase the pressure in the engine bay and reduce cooling airflow across your radiator (the Evo X and SRT-4 were clearly designed to deal with their relatively small turbo-cooling vents).
4: An Evo IX-style hood vent, behind the radiator and above the turbo, can serve double duty of improving airflow across the radiator at speed, and letting turbo heat escape when you're stationary. Ideally, the vent should have a small lip at the leading edge to create a low-pressure area over the vent that encourages air to leave the engine bay.
5: I'm too lazy and too much of a carbon snob to have anything but a stock hood. Some day, when I find a properly vented hood made of real, vacuum-bagged, autoclaved dry carbon, I'll want one. But then it will be too expensive and I'll just drill a hole in my hood instead.
In case you missed it, #4 was your answer.