Project EP3 Civic Si: Fixing EP3 Bumpsteer with Hardrace

by Mike Kojima


In prior editions of Project EP3 Civic Si, we went over the history of the car, the worst handling Civic made with suspension geometry taken right from the Stream minivan. Well, so far we have gotten rid of squishy rubber and added swaybars with help from Whiteline. We also got rid of the flaccid shocks with a set of coilovers from Fortune Auto

Perhaps the worst issue with the suspension of the EP3 is the large amount of bump steer present in its steering geometry. With its origins coming from a  van, apparently, the steering was designed for ease of running a steering column instead of any kind of aspirations of good handling. In fact, the layout of the steering is perhaps the worst that we have seen in a modern car!

Ideally, in a McPherson strut front suspension, the steering rack should be placed so the inner tie rods pivots are in line with the lower control arm pivot bolts with the outer tie rods being in the same plane as the ball joints. On an EP3, the steering rack is midway up the firewall. The rack is also very short and the tie rods very long; this makes for a lot of toe in under suspension moments.

The constant big changes in toe makes the car squirm around and wander which is annoying, and in some cases, it's so bad it makes it hard for the car to stay online.  All is not lost- we have obtained some parts to help reduce the bump steer, hopefully, to manageable levels. 

To read more about Project EP3 check here!

We got a set of these bump steer correcting tie rod ends from Hardrace. These parts relocate the tie rod from the top of the knuckle to the bottom.  This moves the tie rod pivot point down about 2 inches from stock. This is about an inch too far for optimal bump steer reduction, but since it is on the far side of the knuckle, it at least can be moved.

Here you can see how the shank of the Hardrace tie rod end has the taper on the opposite side as normal.  The tie rod end itself is a spherical bearing rather than a ball joint. This makes for less play. The spherical bearing is shielded from the environment by a rubber boot. 

We also installed Hardraces roll center correcting ball joints. The balljoints have a longer body which relocates the pivot point of the of the balljoint almost an inch lower. This raises the roll center and goes a long way to correcting its location which dropped when the car was lowered. Correcting the roll center height increases geometric anti-roll and helps improve the poor camber curve of lowered McPherson strut suspension.  If you want to learn more about roll centers check this out. 

On an EP3, the ball joint presses into the knuckle instead of the control arm so to relocate the pivot of the ball joint, it has a longer body to move the pivot down. Your typical roll center correcting ball joint that presses into the control arm has a longer shank to do the same job. The relocation of the ball joint pivot lower corrects the too low position of the inverted Hardrace tie rod end!


Page 1 of 6 Next Page
Bookmark and Share
Monday, June 12, 2017 10:05 AM
Did you measure bumpsteer before and after? Or is using the old eyeball good enough for most situations?
Mike Kojima
Mike Kojimalink
Monday, June 12, 2017 11:22 AM
To be honest I should have but I wasn't taking this seriously. I didn't think this would make as much difference as it did.
Monday, June 12, 2017 1:07 PM
Man, wish Project EP3 was around a decade sooner. I have had my EP3 for over 10 years now and have racked up over 200k miles. My wife is always the one that gets the new "family" car so I have been using my EP as my daily and have worked on that car from bumper to bumper. It is a fun car especially since it has the complete RSX-S swap with the K20A2 and matching 6 speed and a mild K-Pro tune. The issue has ALWAYS been the suspension though.

The EP3 is a love it or hate it car but after watching them way back when on the BTCC and inspiration from the Top Gear episode where it dusts the FN2 CTR on track (https://www.youtube.com/watch?v=XeGAmcLYGQ8) I have always loved the car and have been optimistic about making my own Type-R replica. Unfortunately I have been through about 4 suspension setups which included 3 different brands of coilovers in which my current setup is the KW V1 Coilovers/ Buddy Club Extended Ball Joints / Progress Rear Sway-Tie Bar Combo thinking that would solve all my issues. It is a mild setup as I also don't "Stance Out" the car with ridiculous amounts of negative camber and stretched tires as I am more on the side of performance over aesthetics and looking "Dope" at the meets (which I never go to). With all that and the decade worth of tuning and maintenance I HAVE NEVER ELIMINATED THE BUMP STEER to the point where I just drove it the last four years thinking that was just the way it is. I had no idea Hard Race offered the Tie Rod ends that allowed mounting under the Macpherson Strut. Though I am probably at the end of tuning the EP3 (It will most likely go to my teenage daughter when she can drive in a couple of years), it has given me hope on finally solving this long term ailment (Like a rash that won't go away LOL). THANKS FOR THE INFO MIKE!
Mike Kojima
Mike Kojimalink
Monday, June 12, 2017 2:53 PM
For sure the ball joints with no tie rods makes bump steer worse. Get the tie rod ends ASAP! I was amazed at the difference and I race a Sentra bump steer machine! Our EP3 turned from the company meah car to one that is super fun to drive.
Monday, June 12, 2017 8:33 PM
Since the tie rod attaches to the coilover shock body not the knuckle, and that is a separate piece of the assembly, why not just cast a new steering arm with the tie rod end in the optimal location? Would seem to be an easy added feature to the coilover kit.
Mike Kojima
Mike Kojimalink
Monday, June 12, 2017 9:38 PM
Some coilover kits have correction built in like you suggest. I think Buddy Club are made like this.
Tuesday, June 13, 2017 9:15 AM
I wonder if there's enough beef to these racks to mount a rack spreader. That, or adapt a proper width rack from some other vehicle.

Then again, there probably isn't enough of a market for this. Does the RSX use the same bullshit steering rack location?
Mike Kojima
Mike Kojimalink
Tuesday, June 13, 2017 9:53 AM
The rack is a stupid short deal and it has numb electric power steering. It's the only car more wrong than a Sentra! With this solution of sorts, it's acceptable and you would only sell like 3 of these if you made something. The RSX sucks too.
Tuesday, June 13, 2017 10:43 AM
The handling made me crazy on this car.
Tuesday, June 13, 2017 11:14 AM
These articles make me miss my Integra GS-R more and more.
Wednesday, June 14, 2017 6:24 AM
Hardrace wants alot for there RC adjusters for Yaris. I was going to modifiy (or make custom) control arms that use american BJs that I can get at the local dirt track sbop. As they make all sorts of BJs. I have Howe low friction stuff in one of my Malibus along with some steering geometry improvements the car turns like a dream.

But that leaves out the tie rod for the Yaris. I do not have much bumpsteer as the rack is mounted low but if i go messing with the control arms then I need to address that too. I wonder if Hardrace would sell just that pin...i bet it would work.
Wednesday, June 14, 2017 5:55 PM
Bump steer is the WORST. I don't know how Honda Engineers went from the Double-Wishbone F/R Civic EK chassis, to the EP3 and still managed to sleep at night.
Wednesday, June 14, 2017 6:02 PM
I thought the rack links were long because they had to be to minimize bump steer (as the link sweeps, the horizontal distance changes less with a long arm, then a short arm). If they were shorter, wouldn't that make it worse? It seems to be a similar situation to SLA suspension.

Maybe, the Honda engineers were trying to 'patch a problem' that they couldn't fix without a complete redesign.
Mike Kojima
Mike Kojimalink
Wednesday, June 14, 2017 6:22 PM
To reduce bumpsteer then the rack is in a really weird position, the inner tie rod should fall on a line from the inner LCA pivot to the top of the strut and tie rods be parallel with the LCA. I am not sure what Honda was thinking when they figured this stuff out. The firewall is really forward so maybe the rack packages better there due to steering column angularity issues?
Wednesday, June 14, 2017 8:33 PM
Different length tie rods and arms make for different arcs. Different arcs make bumpsteer. You can minimize this by moving the link locations, but this is a mitigation, not a solution. Sometimes, you end up with a little bumpsteer because shit gets in the way or as a compromise on some other geometry goal, but this is an absolute disaster.
Wednesday, June 14, 2017 11:17 PM
I agree, Mike. The most interesting question is why Honda chose to design it this way in the first place.

For one, as a previous generation Civic owner (EG6), my number one complaint would be chassis rigidity and interior noise. They most certainly addressed this issue, in the EP3 model. So, out of curiosity, I found a couple of old reviews from early 2001.

Quoting the original Car and Driver Review:

"Like every new Honda, the Civic boasts improved structural rigidity over its predecessor. In this case, according to Honda engineers, it's 95 percent stiffer in torsion and 22 percent more resistant to bending. Furthermore, although the new car weighs a few hundred more pounds than the old Civic, Honda says the use of high-tensile steel saves 44 pounds. Electric power steering further lightens the weight of the car. And, say Honda spokesmen, all the body gaps are smaller, and the fitting tolerances are tighter.

Indeed, the new hatchback feels like a bank vault when on the move, such is its sense of solidity and quietness.

The chassis feels secure and controlled on the road, the hardware is similar to that of other Civics and Acura's new RSX, with struts up front and links in the rear, and the calibrations are, according to chassis specialist Yuichiro Yanagasawa, almost exactly like those of the Euro Type-R Civic. No wonder it feels like a European car."

For giggles, I decided to watch the original Motorweek review on YT(I really enjoy watching old car reviews for some reason):

And Motorweek stated, "MacPherson strut front end gave engineers room to utilize a high-mounted electric steering rack with extra-long control arms for improved toe control and tracking stability."

My questions:
Despite the location of the rack, does the EP3 ALSO have 'worse' bump steer because of electric steering?
Does 'improved toe-control' adequately explain the ridiculously long steering control arms? (it does actually makes some sense, because the longer arm has a wider arc, but the entire high-mounted rack introduces a ton of bump steer, which is the 'opposite' of improved toe control)
Mike Kojima
Mike Kojimalink
Wednesday, June 14, 2017 11:32 PM
I think the long tie rods are a compromise for the weird rack location. I possibly think that the EPS with its switching delay and resulting odd on center feel makes the sensation of bump steer worse. Improved toe control probably means improved over total crap! It's the same stuff I had to deal with when racing Sentra dollar cars, those have almost as much bump steer as wheel travel. It also goes to show you that a lot of (not all) automotive journalists don't know what they are talking about! A new problem that has started to crop up in the last few days is that since the car got fun to drive and the front has way more grip, everyone is railing the car around and the additional force on the odd center linking tie rod bracket is making the steering rack develop an up and down twisting play which generates its own new type of bump steer which causes the car to be a little scary. We have to figure out how to deal with this issue now. Ugh.
Thursday, June 15, 2017 11:09 AM
Yeah, but the good journalists do command a 'cult-like' following. Thank, Mike!
Thursday, June 15, 2017 5:15 PM
The long arms make no sense. The bigger the delta in control arm to tie rod length, the worse bump steer will be.
Mike Kojima
Mike Kojimalink
Thursday, June 15, 2017 5:38 PM
They make the car toe in under roll.
Friday, June 16, 2017 9:48 AM
3 degrees of negative camber! That seems beyond streetable. What would you deem to be ideal for the street, with the end goal of balancing cornering grip with even tire wear? I've found that ~-1.5 with no toe works across a wide range of suspension types.
Mike Kojima
Mike Kojimalink
Friday, June 16, 2017 4:51 PM
that's what you run, this is what I run on most strut cars.
Friday, June 16, 2017 7:29 PM
I will second the 3 degree of Negative Camber on struts. I dialed in approx 3 degrees or so on the front of a VW Jetta, and the front-end is solid in the curves, where it felt very sketchy before. I have about 1.5 degrees of NC on the rear, too.

Strut equipped cars absolutely need more static camber. Even more so, if they are already lowered.
Friday, June 16, 2017 8:44 PM
If you drive like a maniac everywhere, tires actually last longer at 3*... :)
Mike Kojima
Mike Kojimalink
Friday, June 16, 2017 10:27 PM
or if you do occasional track days, much longer.
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2017 MotoIQ.com