Project subaru WRX STI

Project Subaru WRX STi Part 4: Brake System Upgrades

By Mike Kojima

The old engineers joke that “speed doesn’t kill, rapid deceleration does” applies to all vehicles capable of attaining dangerous velocities.  Since we have been concentrating on improving project STi’s acceleration and handling, we have definitely achieved the ability to reach deadly velocity on the track.  In our first foray onto the grounds of Willow Springs raceway we noted that perhaps we were starting to drive beyond the capabilities of the stock brakes due to the higher speeds that we were able to attain on the straights.

To read the rest about Project Subaru WRX STI click here!

Project Subaru WRX STI
Since our STI gets driven on the track, we had to do some brake upgrading.

The STi is blessed with excellent, large, competition worthy Brembo brakes from the factory.  However these Brembos are somewhat hamstrung by the requirements that the STi be a civilized street car.  Street drivers usually don’t tolerate any sorts of squeaks and squeals that high performance brake pads are prone to.  Thus the brake pads that the car is equipped with from the factory are compromised to be less than optimal in hot braking performance with quietness and effectiveness when cold.  Usually a brake pad works the best within a given temperature range.  A pad that works best when hot, often works poorly when cold and vice versa and Subaru compromised in the conservative cold direction.

Project Subaru WRX STI
When comparing the stock rotor to the DBA rotor, the weight saving alloy hat and slotting are easily noticeable.

When driven on the track the stock brakes were not horrible as stock brakes often fade away to nothing when pushed hard.  We did notice that the brakes did fade, losing about 25% of their effectiveness then stabilizing for the duration of the track session.  Although this was not horrible for OEM brakes, it was not exactly confidence inspiring either. Although we could have simply installed a super duper mega huge brake kit, we decided that we did not really want this approach for several reasons.

Project Subaru WRX STI
When viewed from the edge, the thermo paint for temperature indication in use and the more open vaning of the DBA rotor becomes apparent.

The first reason is that the 2004 STi is prone to suffer what is called brake caliper piston knockback.  The 2004 tends to wear its wheel bearings.  This causes the wheel and brake rotor to wobble around when subjected to cornering load.  This wobble knocks the brake caliper pistons back into the caliper.  When the brakes are reapplied, the brake pedal height is now lower than usual because the pistons have been knocked back into the caliper and must be pushed back out to contact the rotor.  In extreme cases the pedal may drop to the floor and must be pumped to regain braking power.  Subaru recognized this problem and the 2005 and later models have bigger hubs and wheel bearings to eliminate the issue.

Page 1 of 5 Next Page
Bookmark and Share
Thursday, October 21, 2010 8:56 PM
This just reminds me again why I wish I could fit a two piece rotor on my Wilwood fronts inside a 15" wheel. Light, heat transfer, all the good reasons why it's a better setup even in a low powered car like my B13...nice article. Rings very true based on a good friend's experience with his own '04 and tracking it, hub upgrade cured many problems.
Friday, October 22, 2010 4:49 AM
It's been on my to-do list for quite some time, which means it likely won't ever happen, but I've been waiting to see some actual measured data regarding F + R strut braces on *these* cars.

Tower-to-tower flex should be measureable with the proper sensors (strain/pressure guage thingy).

It would be a great "SIGNAL" in the "More Signal, Less Noise" motif, though I understand a site like this is largely no longer independently in a position to publish data that could (or not!) ruin relationships with sponsors. Ahem.
Mike Kojima
Mike Kojimalink
Friday, October 22, 2010 5:08 AM
If you would like to take the time to come here with your equipment and do the test, we will be glad to publish the results. Or if you would be willing to sponsor us with equipment we would be glad to do the test you desire.

How much did you pay for your issue of MotoIQ?
Friday, October 22, 2010 7:38 AM
I'm sure towers flex in all directions, but do they usually flex inwards or outwards?

How about attaching a rod to each tower and rig it so if the flex one of the rods moves a zip tie on the other rod. You could do 2 different zip ties, one that moves if they flex inward and one that moves if they flex outward. That would give you the peak flex in either direction rather than a log of the flex over time.

Sounds cheap and easy to me. I'm not sure it would work, but it might be worth a try if your interested in such data.
Der Bruce
Der Brucelink
Friday, October 22, 2010 7:58 AM
Well, I've been paying free.99 for MotoIQ for a while, and must say this article was extremely well written and I've never really noticed too many biases in prior articles, if any.

Alright, Eric does seem to have some strong opinions about engine management!

Any idea on body roll improvement with the strut towers vs say bigger adjustbale sway bars or just improved chasis rigidity? Either way, I'd imagine the increased responsiveness was nice.

I can only speak from personal off-roading experience but the SS brake lines are one of THE most important brake upgrades you can make to improve pedal feel and avoid fade. Stopping a heavy 33-35 inch wheel/tire combo anywhere has serious consequences, namely brake line swell, so I can only imagine how having to bringing a 3300lb car from 80 or 100 down to 30mph in 70ft, REPEATEDLY, must be on the lines.

Any idea on how the Motul would hold up in my Pacific NW winter temps?
Mike Kojima
Mike Kojimalink
Friday, October 22, 2010 8:10 AM
Burm, no it won't work.


Strut tower bars have no affect on body roll, they reduce chassis flex which in turn usually improves ride as the chassis natural frequency goes up and improves sensitivity to chassis adjustment.

Motul is actually not as bad for water absorption as other racing fluids like Castrol SRF. It has a pretty high wet boiling point.
Friday, October 22, 2010 8:38 AM
Since we're talking about mechanical engineering on chassis bracing, I've always had a question on those.

Most are built with heim joint ends, which makes for easier bolt-up and adjustability.

Of course, it also sets up the bar to instantly be in bending and not transmit any moment to the mounting plate.

Wouldn't you end up with a way stiffer brace if it didn't have a joint between the mount and the bar? e.g. if it were a bar rigidly bolted with like 4 bolts to the mounting plate?

With a joint, the beam's automatically going into buckilng, whereas with a rigid mounting plate, you have shear working for you.
Der Bruce
Der Brucelink
Friday, October 22, 2010 9:29 AM
I didn't get my degree in mechanical engineering but a lot of that "bending" your referring to would probably come from poor material rather than poor design.

A high thread count screw threaded into the bar properly is kind of like a sleave properly welded into tube, they act as one with possible flex measured on the micron level. A well designed heim joint should also translate minimal movement or flex. The mounting plates on some of those e-bay types due tend to come with thin mounting plates which should be one of the biggest concerns.

These designs are, however, as you said, for "easier bolt-up and adjustability."
Friday, October 22, 2010 10:56 AM
Upgrading my hubs to the 05+ spec made a night and day difference for my car at the track. Did the pads, upgraded fluid, and SS lines, but nothing made more of a difference than the stronger hubs. It's not the cheapest way, but it stops the knockback problem at its source. Plus you get a better selection of track wheels with the 5x114.3 bolt pattern. Another item which helped was a master cylinder brace, the stock one flexes under hard braking.
Friday, October 22, 2010 12:22 PM
@ Albert: What kind of costs ballpark would one be looking at in upgrading from the '04 hubs to the '05+ spec hubs/hardware?

Prior to this article I was unaware of the hub diferences between an '04 and an '05. Would you say I'd be better off if I was shopping for an STi to limit my search to the 05+?
Friday, October 22, 2010 1:12 PM
I would not say the Motul 600 is comparable to the SRF or the AP fluid.

I have used Motul 600 in the R34 GT-R World Challenge race car. We would boil it nearly every session. It came out of the rotors looking pretty bad. Being the stubborn person I am, I refused to spend the money on the SRF. I was totally skeptical to the SRF being that much better than the Motul. On paper, they looked about the same. However once I changed the Motul for SRF, I learned that there are more than just wet and dry boiling points to consider.

Once we changed to Castrol SRF, we hardly got any bubbles out of the calipers. The pedal didn't get mushy, and when we did bleed it, the fluid looked much better than the Motul.

In the Scion, one car would need a bleed about every race. The other car could go 3-4 races.

The Motul is a good street fluid, fine for a lighter car, but I wouldn't use it in a serious heavy track car.

Friday, October 22, 2010 1:21 PM
"It came out of the rotors looking pretty bad."

Or, it actually came out of the calipers looking pretty bad. If it came out of the rotors, I imagine things would have been worse.

No easy editing here.
Mike Kojima
Mike Kojimalink
Friday, October 22, 2010 5:31 PM
I never have had any problems with Motul even under the hardest endurance racing conditions and have found Motul to last a whole season or as long as the caliper without bleeding! I have found SRF to not last as long but to be good if you bled frequently.

In the scope of my experiance, SRF is not worth it because I have never gotten fluid fade with Motul ever, even when the brake temp exceeded 1100 F. Even with old Motul.

I would say you have more experience with underbraked cars and racing though. All of the really fast and heavy cars I have worked on had properly sized brakes. We used 600 on this car because I have a huge stock of it. We will be switching to 660 soon on all of our project cars.
Friday, October 22, 2010 5:49 PM
There are couple of pictures showing the removal of the brake shields, but nothing about those in the article text beyond the pic captions.

So ... what's the purpose of those shields and why do all OEMs put them on the cars. I guess it is not for channeling cooling air if you say the rotors cool better with those removed?
Friday, October 22, 2010 6:13 PM
I have never tried the 660. Just the Motul 600. As I said, I think its fine on a lighter street car, street pad setup.

However, I don't think its adequate for a Nissan GT-R. I think it might be marginal on a STI that is run with a fast driver. It was not good enough on a World Challenge Skyline GT-R.

1100 F = 593 C. R35 guys regularly see 650 C - 700+ C. They have a large rotor - 380 mm , 6 piston calipers (front). No ducting or diverters stock. Run the car hard at the track, and you will crack the factory Brembo rotors. Some guys can do it in 1-2 hard track days. The guys that are running the R35's hard, are getting mushy pedals with Motul 600.

I will take the SRF at $70 a bottle, or the AP PRF at $29 a bottle for not having to have the pedal sink to the floor at the end of a 150 mph braking zone. I will gladly replace it more often.
Mike Kojima
Mike Kojimalink
Friday, October 22, 2010 8:17 PM
The shields are dust and water shields. Some cars like EVO's have air scoops in them but they get in the way of bigger rotors.
Saturday, October 23, 2010 3:16 AM
Great series of articles Mike! Did you happen to weight the DBA5000 rotors? I'm curious if you saw any weight savings.
Dusty Duster
Dusty Dusterlink
Saturday, October 23, 2010 3:22 AM
I noticed on my car that rear chassis bracing made a bigger difference than front bracing. My guess as to why is that there's an engine up front (which isn't very flexible), but there is nothing in the back of the vehicle other than trunk space. This was most definitely the case on my car, which is a hatchback.
Saturday, October 23, 2010 4:43 AM
On the shields - if those are to keep water away, why no such shields on the outer side? The rotor is usually in the middle of the wheel, so with alloy wheels water can enter on the outers side just like on the inside?

About the dust - which direction? - is that the brake dust off the pads or dust from the road?
Mike Kojima
Mike Kojimalink
Saturday, October 23, 2010 6:22 AM
I think its to keep those little rocks from getting between the rotor and caliper. Part of most OEM durability testing is extended driving on a gravely surface. I also think its to prevent corrosive brake dust from going all over the suspension. I have been told that pad wear can increase with them off. Some cars like the EVO have cooling scoops built into them. I think my Z32 did as well. Probably if the shields were on both sides the brakes would get way too hot? They sell aftermarket dust shields that fit on the wheel side to keep your wheels clean but there are warnings about brake cooling with them.
Mike Kojima
Mike Kojimalink
Saturday, October 23, 2010 6:24 AM
Shift, I did weigh them but I forgot where my notes were. I know thats bad, I would like to say they were something like 3 lbs per side lighter
Sunday, October 24, 2010 3:44 AM
The conversion isn't too hard if you have a later build date 04 like I did. Need the 05+ front axles, front knuckles and the hubs, all perfect OE bolt on fitment. If you are looking to spend a lot of time on the track, I would recommend going with the 05+ so you don't have to deal with this at all. The 04 probably experiences less knockback with fresher wheel bearings, so I guess if you don't mind replacing them every season, the 04 is fine too.
Sunday, October 24, 2010 8:19 AM
Thanks Albert, knowing what's involved now certainly makes me think that it would just be a lot easier to go with an '05 right off the bat. I don't think I would save enough on the purchase of an '04 versus an '05 to justify swapping over all the pieces.
Mike Kojima
Mike Kojimalink
Monday, October 25, 2010 9:06 AM

Do you know what year that they cheapend out things and went to steel control arms instead of forged aluminum? Was that in 2006?
Tuesday, October 26, 2010 6:52 AM
Albert, thanks for the reminder, a MC brace is on my to-do list for the G, too. There's quite a bit of movement with the current MC, not that I'm complaining about the feel or effectiveness of its Brembo's. Pads and rotors had same issues as your STi, and fluid was replaced with Motul awhile back. OEM would not have survived track use long.

Great answer Mike K re:bracing STB F/R, of course does nothing to reduce body roll, but makes all the difference in feel and effect of other adjustments and suspension mods. Granted I start with a noodle in a B13, but the rear STB with triangulated bracing down into the trunk floor was an eye opener. I never realized how much slop there was back there even with the "IKEA" brace:

And, IIRC yes Dave C already did do measurement and testing with his homebrew front triangulated brace regarding chassis flex, it had something to do with using a laser, mirror, etc. I mostly nodded a lot hearing what he did, so pls excuse the lack of details. Once the front triangle was loose at the FW. I felt immediately like something was seriously wrong with the car. A thicker backing plate, loctite and re-torque of bolts, back in biz. Note the FSTB is just a box section of steel stock, it works great (and no sponsor issues, jeez what a bunch of crap that line was, as if MotoIQ has to suck up to its sponsors...) -
Tuesday, October 26, 2010 7:00 AM
@Der Bruce - see above rear bracing pics, that is *exactly* what that builder does with the heim joints, uses top notch yet light materials, high thread count (which btw being threaded, I can get an exact fit vs. 'close' with welded construction). The lower mounting plate, bolts, and L brackets connecting from the trunk to the RSTB are very beefy, and the lower plate uses holes already reinforced in the chassis for the gas tank brackets. All the hardware is very strong, yet overall very light package. A 1-piece welded FSTB triangulated is a serious PITA and heavier, too. You're also not allowed to weld bracing to the chassis in many ST/SP classes outside of cage-related stuff.
Wednesday, October 27, 2010 3:20 AM
Steve: "as if MotoIQ has to suck up to its sponsors" Yes, as if. Plenty of the articles are advertising. That's just how this game goes for the most part. Magazines, etc. If you don't get that, we probably don't have a lot to discuss.

Burninator: I don't know if yours would work specifically, but there surely are many ways to perform the test with some ingenuity, time, and little to no cash.

Mike K: I don't really know what to say. It was a suggestion/idea. Even that apparently offended you.
Mike Kojima
Mike Kojimalink
Wednesday, October 27, 2010 5:21 AM

The articles are not advertising, a lot of the parts are purchased, probably 80 percent of them at this point. We select the parts we want to test, we avoid the parts we know are not going to work. Then we evaluate them to the best of our budget which isn't very much. We are currently a non profit organization.

Like I said if you don't like the way we test something and if you are willing to donate your time and equipment to help us test to your liking, please come down here and help us out. As long as your testing methodology is sound, we will publish the results. If you have a business model for a way we can fund this place with no advertising revenue I am all ears as well.

Peanut gallery comments are lame and is offensive, helping out isn't. If you think we suck that bad and have such a negative attitude toward what we do, please don't come here anymore. You clearly think our tech and stuff is lame and beneath you, why bother, you know more than we do and this place doesn't benefit you.
Wednesday, October 27, 2010 10:23 AM
Mike: Far from it -- I am sure you, Eric, and Dave know twice as much as I do related to auto tech/engineering. I thought it was a fine article, but one that has been written 10 times before in the last 5 years, and I have always thought this site was pointedly different in its goal. I had really high hopes for MotoIQ when I originally saw the content being generated. "Finally, someplace with some depth. We don't need to learn how to install brake pads anymore. That exists elsewhere."

So, perhaps my standards or expectations for MotoIQ are too high (that is NOT a jab, I am speaking honestly) and I need to just lower them (or as you suggest, not come around anymore).

I'd be happy to PayPal the crew $20 for the detailed/deep articles I've enjoyed in the past. Just give me someone's PayPal address.

I'd also be happy to pay an annual subscription to any site performing independent testing (with data acquisition) or articles with good tech ideas/findings/thoughts, etc.

Have you ever considered taking my comments as motivation instead of as personal attacks? I think you, and sure Eric, have labeled me in your heads as "some hater", and that's not the case at all. Maybe we just see the world differently. Maybe it's cultural, or coastal, or who knows. All I know is, I am psyched to get ANY feedback on my work, positive or negative. It is a gift of response from an audience. If someone says, "This article sucks ass.", I'm going to respond with, "What didn't you like about it?" and *HOPE* they respond, because I'll benefit from a different point of view.

So ah.. yeah. PayPal address?
Mike Kojima
Mike Kojimalink
Wednesday, October 27, 2010 12:25 PM
Perhaps its the sarcastic intonation of your posts that we find irritating. Its obvious that you are a smart guy. Of course we would like to do more intensive testing but we don't have the money to buy the equipment and track time to do it. You have got to realize that track time costs several thousands of dollars per day and even the new cheaper forms of general data acquisition costs several thousands of dollars as well.

Not to mention the time involved. We already work 12-16 hours a day. We can't afford to hire another person yet. I don't mind criticism and disagreement but comments like "It would be a great "SIGNAL" in the "More Signal, Less Noise" motif, though I understand a site like this is largely no longer independently in a position to publish data that could (or not!) ruin relationships with sponsors. Ahem. " are pretty insulting.

Think about our position, we work our ass off to make this the best tech site on the web which is a FREE resource and to read shit like that pointed at us when we are bone tired from staying up all night to get stuff done, well no duh we are grumpy about it.
Eric Hsu
Eric Hsulink
Wednesday, October 27, 2010 1:26 PM
Jeff, I don't think you are "some hater". I do think your responses reflect those of a smart ass however and that really isn't necessary. Just like you think that Mike should respond to your remarks in a different way, we believe you could express your comments in a different way. Instead of being a smart ass and possibly offending somebody in the process by clowning the site's concept/motif, you could just say "Dude, I think your brake story is too broad. Can you guys post some test results next time?"

Remember that its easy to offend somebody who puts countless hours of hard work and personal fortune into something. Even if you think MotoIQ isn't up to your standards as far as quality of information goes, you'd have to admit that the bulk of the information is significantly different and written by those who actually have experience.

Then again perhaps you're just being yourself which is fine. I'm not going to tell you to cater or change your personality to MotoIQ. But I will tell you that smart ass remarks are not appreciated by anybody who pours their hard work and money into something only to have it belittled by smart ass comments. I think that's going to be the case pretty much across the board anywhere you go be it cultural, coastal, or who knows differences or whatever.
Friday, October 29, 2010 8:13 AM
This exchange has made me pause enough to more carefully consider where the problem is.

You're right. I've been dickish for no reason sometimes here. I apologize.
Mike Kojima
Mike Kojimalink
Friday, October 29, 2010 8:19 AM
Don't worry about it!
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com