Honda Insight Forum banner

1 - 20 of 36 Posts

·
Premium Member
Joined
·
308 Posts
Discussion Starter #1 (Edited)
Rear sway bar open source brainstorming thread

We've waited a long time for a rear sway bar to materialize for the G1 Insight but for one reason or another, not much has happened on this front. My goal is to drum up some interest in open-sourcing this project to see if collectively we can experiment, iterate and hopefully settle on one or more designs that work.

Background:
I'm used to driving cars that handle better than the stock 1st Gen Insight (e.g. Subaru WRX STi, Lotus Elise, Honda Fit Sport, Nissan Sentra SE-R, etc.) but I believe Honda left a lot of potential on the table when they built this 1856 lb car w/suspension design similar to a Fit, Golf, Rabbit, etc. So my goal is to improve upon areas in the chassis setup that Honda might have optimized for a handling aesthetic different than my own. After seeing what it feels like to do the following mods:
New OEM shocks/struts & GAZ adjustable rear shocks, stock, Aveo & Matiz rear springs, 25mm rear spacers/adapters, and so on, I think the rear axle beam wasn't designed to be stiff enough to adequately limit body roll and thus makes the car feel less nimble than it should for its weight. I know that Guillermo adapted a Shine rear sway bar designed for a Golf/Jetta a long while back and said it transformed the car but being designed for a much heavier car was too extreme. So I'd like to help design and build a way to stiffen the rear torsion beam that's adjustable or at least can be altered to suit different tastes, conditions, etc.

Goals:
--Increase rear torsion beam stiffness to improve transitional response (e.g. make the car more eager to turn-in and roll less)
--Bolt on installation without drilling any holes (Shine's VW rear sway bar requires drilling multiple holes with a cobalt bit)
--Staged or adjustable stiffness
--Lightweight design

Ideas (note: I haven't fabricated any of these yet but am hoping to inspire someone with fabrication skills/connections)

#1 - U-bolt/muffler clamp using square tubing (drawings are in progress, crude sketch to start)
This design clamps onto the beam like an exhaust clamp and uses all-thread bent into a U-shape and a section of square tubing across the open side of the axle beam. By boxing in the open C-shaped beam profile, you stiffen it. By using two of these clamps, you are stiffening both sides of the beam equally. The square tubing section is notched and drilled so that a section of square tubing can be added to span between the two clamps and serve as an additional torsion bar. The wall thickness of the tubing can be varied to adjust stiffness further. Since the axle beam is already spring steel, my design assumes steel threaded rod and steel square tubing for high strength and no galvanic corrosion issues.

#2 - a bent plate variation on the above instead of square tube (square tubing design is easier and probably stronger/lighter though)

(I have more ideas but I think #1 is the most feasible and I need to sign off here)

Who is with me on this adventure? :)
 

Attachments

·
Administrator
Joined
·
12,148 Posts
IIRC Paul Andrews and the Oaktech UK G1 Rally car had a rear sway /anti roll bar of some sort fitted.
I can't remember if it was bolted on or clamped on.

I'll e-mail him.
 

·
Registered
Joined
·
50 Posts
Rear anti roll bar

Hi all, yes I built a rear anti roll(sway) bar into the back of the rally car back in 2006. It made a big difference to the nimbleness and turn-in performance of the car and gained us a huge advantage in stage times compared to the standard set up. I also use 30mm rear wheel spacers, shortened Matiz springs and bespoke dampers with remote gas bottles.

I went the route of drilling the beam. My ARB has 2 posts at each end which are a tight fit inside the beam. I fabricated it from a length of steel tube. The first beam I built fatigued by the inner post on the drivers side after 7 or 8 rallies so I made a new one and sleaved over the ends round the posts to add strength. It has been great ever since....we nip the bolts tight before every rally.

These mods are great on a track with a CVT car. You chuck it into a corner and balance it by throttle & left foot braking to keep the nose of the car pointing exactly where you want to be going to keep maximum speed through a corner.
Lowering the rear of the car makes a big difference to controlling high speed oversteer but the biggest bugbear to this is the lack of travel of the springs and dampers. The ride on my car is really bad on rough surfaces!
My next mod would be to build some damper mounts up inside the rear wheel arches and use much longer travel dampers. These will fit with spacers on the back wheels.
 

·
Registered
Joined
·
41 Posts
It seems a good starting point would be to know the current situation.

Coming up with a good rear swaybar will be a balancing act between front and rear roll stiffness.

Since the rear suspension does provide some torsional resistance, it would be good to know what that is.

Same for the front. Does anyone have the geometry of the front swaybar? A minimum would be the diameter, width from side to side, and length of the arms. From some rough measurements we could get a reasonable number for the front roll stiffness.

Of course, it would also be possible to simply measure it - front & rear.
 

·
Registered
Joined
·
41 Posts
Perhaps I'm a bit slow here, but I'm trying to get my head around the problem.

If I state anything incorrect here PLEASE correct me.

In general, it seems lateral load transfer to the outside is bad, as it reduces traction the inside tire could otherwise provide.

There are several ways to do this, according to Carroll Smith:
1) Increase track width - - can be done with wheel spacers
2) reduce weight - - difficult on an Insight
3) lower CG - - can be done a little bit by lowering

So, reducing body roll doesn't have a large affect on lateral load transfer, but it can adversely affect suspension geometry. Also the time taken to settle the suspension is important, so it should be minimized.

To increase roll resistance, there are these possibilities:
1) Increase suspension stiffness (springs or "dynamic springs" - dampers)
2) add or increase stiffness of anti-roll bars
3) raise the roll center to reduce the roll moment - - Pretty difficult with the Insight

I've measured the front anti-roll bar and found (ignoring all the wiggles) a length of 32.5" between pivots and an arm length of 11". The diameter I measured was 0.680", which is fairly ***inaccurate*** because it includes the powder coating.

This produces a torsion spring rate of 125lb/in of deflection, which is probably lower in reality.

I've been playing with the dynamic calculator on the Autocross to Win site:
Autocross to Win (DGs Autocross Secrets) - Dynamics Calculator

FWIW, I'm using a rear spring rate of 106lbs/in for the rear and 140lbs/in for the front. I have no idea what the rear torsional stiffness is, so I'm using a very small number smaller than the front.

By putting in the data of the stock Insight, you can prove to yourself that there is an understeer condition; nothing new. Because the spring/damping is also tied to the roll resistance, it's possible (probably not the best solution) to choose a front/rear spring rate combo that pushes the car toward oversteer. For example, increasing the spring rate in the rear to 150lbs/in without changing anything else pushes the car toward oversteer and reduces the body roll a small amount.

Because the front suspension of the stock Insight is so soft and the front roll stiffness is also soft, it would require a relatively stiff rear roll bar to push the car toward oversteer.

On the other hand, if the front and rear suspension is stiffened with higher spring rates and well matched dampers, it would be extremely easy to push the car toward an oversteer condition by increasing the rear roll stiffness a small amount.

So if the goal is to reduce body roll, it seems that the actual "best" rear swaybar stiffness will be tied to both the front and rear spring rates AND the FRONT roll bar stiffness.

Thoughts?
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #6
I'm on my phone so my response will be a bit brief. I'm going to look into the calculators, etc that you've provided and appreciate the more engineering-oriented approach to improving the car's handling than my less sophisticated empirical/iterative approach.

My goal in terms of handling for the Insight is to make it feel as light as it actually is with a more crisp response to driver inputs. I don't really expect to make it "faster" through the twisties but I definitely want it to be more engaging. As it is right now, with Matiz rear springs, GAZ shocks & 25mm spacers I find the Insight feels heavier than my 2010 Fit which actually weighs ~700 lbs more. I assume that the rear torsion beam stiffness has a lot to do with that plus relatively decent springs & damping from the factory, quicker steering and other refinements that were never a priority on the Insight.

From my empirical testing of 3 different rear spring rates including OEM, Aveo & Matiz, the stiffer springs did not improve transitional response or a general feeling of "nimbleness" by much.
 

·
Registered
Joined
·
41 Posts
Sweet! It seems we have the same goals :)

BTW, if you don't have it already - Carroll Smith's "Tune to Win" has excellent info about suspension set-up.

Maybe mine is the semi-cheapscate approach. I don't mind spending the money on getting good parts - I just don't want to spend a bunch of extra money getting good parts that are wrong for the application, then buying more new good parts because the first ones didn't do what I hoped they would. I recognize that we are forging new ground for the Insight, but its suspension really is not that new.

I'm not saying this method is better. And the number crunching certainly would have more value when tempered with real world experience. Your experience with different suspension bits will definitely help provide information to find a good solution.

What were the spring rates of the Matiz and Aveo springs? Do you have damping rates (preferably the full curve) for the GAZ shocks? Did you change the front?

By "heaviness" are you referring to body roll and understeer?

Thanks for helping me understand this,

E*clipse
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #8 (Edited)
@E*clipse: Glad to hear we share a good amount of common ground in this regard. Speaking of books, I have the Carroll Smith book, the Fred Puhn book and at least a dozen others that cover suspension design, behavior, modification, etc. The net result of a lot of reading and thinking about such things is that we need more data (like spring rates, damping curves, lever lengths, corner weights, roll center heights, etc.) and/or more empirical data from experiments to see what works.

Knowing that this worked but was too extreme for most folks/conditions: http://www.insightcentral.net/forums/modifications-technical-issues/9770-rear-sway-bar-installed-insight-works.html

I was inspired, as were others, to try something similar but more tame.

Spring Rates: (MODERATORS: I think a sticky thread w/this info would be great after mucho searching to dig these up)
Stock: ~90lb/in rear, ~111lb/in front
Matiz: ~123 lbs/in rear, N/A front
Aveo: dunno exactly but I think they're stiffer than the Matiz. Here are dimensions: http://www.insightcentral.net/forums/modifications-technical-issues/16601-aveo-spring-mod-w-pics-9.html#post256794

Miata: (cut to 9.5") ~150lbs/in rear

This post from EQ1 has some great info:
http://www.insightcentral.net/forums/modifications-technical-issues/19842-installed-matiz-rear-springs-over-week-end-16.html#post231527

Regarding damping, I have zero data on any damping rate curves. My fronts are still OEM Honda units I replaced about 5 years/60-70k miles ago (hmm?).

Lastly, regarding my definition of "heaviness": I mean mostly the level of response to driver inputs, especially transitional response through a slalom, on a curvy road, dodging little Johnny when he darts after his ball, etc. If I turn the wheel and the car goes through a sequence of motions *other* then moving where I've pointed it, then I call it "heavy feeling". Antonyms: Agile, Eager to turn-in, Responsive, Nimble, Lotus-like :)
 

·
Registered
Joined
·
41 Posts
A "Lotus - like" Insight would be awesome! :)

So reducing body roll with stiffer springs and beefed up swaybars would be a start. After that getting rid of the mooshy bushings would help out too.

I really like the idea of a sticky with some spring / damper data along with information about how changes affect the ride and responsiveness.

In the next couple of weeks, I'm going to measure the exact corner weights of my 2003 CVT.

I'm also going to measure the stock spring rates. It seems there may be a discrepancy for the rear spring rates (90lbs/in > 106lbs/in). This could be because of the different vehicle weights, fatigue, or measurement error. We also don't have any numbers for the front. I have a pair of front struts on order from a CVT insight. I'll measure both spring rates to get a data point and check consistency.

While I'm at it I'll measure the front sway bar and rear swingarm spring rates.

It seems eq1 had good luck with a 150lb/in spring in the rear. He also found that 216lb/in was too stiff and the stock shock couldn't control them.
 

·
Administrator
Joined
·
12,148 Posts
The front springs in the Manual and CVT are different due to the extra weight of the CVT gearbox. So you might need to measure both types.

Rears are the same for both cars ASFAIK.
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #11
Lotus-like is definitely my goal--my other car *is* an Elise and weights 100 lbs MORE than the Insight :). Of course the two cars were designed to meet very different goals but they share lightness and I'm curious to see what it would take to bring the Insight a little closer to the driving feel that makes the Lotus so magical. To that end, I've had a set of front control arms rebuilt to use spherical bearings in place of rubber bushings http://www.insightcentral.net/forums/modifications-technical-issues/36025-experiment-replacing-lower-control-arm-bushings-spherical-bearings.html, have a spare steering rack ready to modify to eliminate the electric power steering like Wocketman did: http://www.insightcentral.net/forums/modifications-technical-issues/24316-removing-friction-steering-system-7.html#post303049, plan on having custom sleeves machined to allow use of VW rear axle bushes like Biv from Russia did: http://www.insightcentral.net/forums/modifications-technical-issues/17742-rear-axle-poly-bushes-8.html#post206178

I'm still trying to stay focused on prototyping my bolt-on rear "sway bar" with my precious spare time as I still think this mod will have a dramatic effect on handling response.
 

·
Read Only
Joined
·
2,009 Posts
I wonder how "Lotus-like" you'll be able to reach considering that your Elise is a mid-engine rear-wheel drive car with a much lower moment of inertia compared with your very front-heavy front-wheel drive Insight. Of the several sports cars I've owned including front-heavy Sting Rays and rear-heavy Porsche 356's, none handled nearly as well as my mid-engined 1970 Porsche 914-6. It's no accident that the best-handling road racing cars are mid-engined and rear-wheel drive.
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #13
@aisbell - LOL, my expectations are a little more realistic than you might think. :) What I mean by Lotus-like is the removal of "slack" and squishy-ness in several areas on the Insight. Honestly, I'd be thrilled if the Insight were more Fit-like. I have no expectation that the anemic FWD commuter car with narrow LRR tires & comparably crude suspension design will ever actually drive like the purpose-built mid-engine track toy--but I do like the idea of taking inspiration from the Elise and applying it to the Insight.
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #14
I made some progress on my bolt-on rear torsion beam stiffening design. It's basically designed like a muffler clamp but is custom built to fit the Insight's rear beam and is meant to be used in pairs. If time/energy allow, I'll get started on #2 tonight. Either way, there's a good amount of snow on the roads in Redmond/Seattle today so I can't road test the design to see if it makes any positive difference. My plan is to try out a pair of these clamps to see how much additional stiffness they provide. If it's not enough for my taste, I'll add the feature mentioned in my original post to connect the two clamps with a length of square tubing that acts as a torsion bar.

Inspiration for this design came from Julian Edgar's work as described here:
AutoSpeed - Anti-Roll Bars and Torsion Beam Rear Suspensions, Part 2
 

Attachments

·
Registered
Joined
·
41 Posts
JoelR,

The "muffler clamp" concept looks very interesting.
Would it be possible to do a few "tuning" tests before putting the swingarm back on the car?
1) Torsional stiffness without the mod
2) Torsional stiffness with the mod - maybe with the clamps in different locations relative to center.

This baseline info could be very valuable as we tweak the Insights suspension. I really like the clamp idea from the adjustability perspective. If for example, the rear end slides too easily, it may be very easy to adjust by moving the clamp positions.

I got the front struts today. I'm going to take a bunch of measurements and post them here. If there's any specific measurement you'd like, let me know.
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #16
@e*clipse: Actually the axle is a spare I bought so yes, tests can be done off the car. :) Any suggestions for how to mount it up & how to measure it? My thought was to use the built-in body mounts to bolt the axle to something very rigid, to prop up one trailing arm into a horizontal-ish position (like it is on the car) and then add weight to the other side to see how much it deflects per qty of weight.

Since my last post, I built the other clamp as well so now I have a pair. Stay tuned for test results!

Joel
 

·
Registered
Joined
·
41 Posts
JoelR,

Excellent! :) I don't have any better suggestion for testing it.

I ran into a snag when taking the struts apart. The top nut doesn't budge. I guess the trick is to use an impact wrench, unless someone has another suggestion.

On another note, I noticed in "Tune to Win" there was a drawing for a driver adjustable swaybar. Basically the arms had a clamp-on mount point for the connecting link. By moving the mount point, the effective arm length could be changed. Shorter arm- stiffer effective sway bar. Perhaps it would be possible to make something like this to shorten the front sway bar's arm? The connecting link is pretty long; it might be possible to move the mount point a little bit with out messing up other geometry.
 

·
Premium Member
Joined
·
308 Posts
Discussion Starter #18
Update on the muffler clamp design: I set up a rig to mount the rear beam to my butcher block workbench close to how it mounts to the car. I then put a jack stand under one spring seat/trailing arm and applied weight to the other trailing arm so I could measure the displacement per unit of weight.

I found that the muffler clamps made no difference with 81lbs of weight applied to one side. Thinking the reason was that the beam distorts less at the outer portion, I moved the clamps inward where the beam appears to distort more. To my disappointment, his made no difference either.

Not to be deterred, I decided to use the remaining piece of 1" square steel tubing I had left over as a torsion bar and bolted it to the muffler clamps. This made a difference! My calculations: @ 81lbs I measured 2 inches of displacement stock (and with the muffler clamps alone) for a rate of 40.5 lb/in. With the added square tube spanning the 2 clamps, I measured 1.375" of displacement or 58.9 lb/in. This equates to a 45% increase in torsional stiffness.

Up next is to mount the contraption to my car & do a road test to see how it feels.
 

Attachments

·
Premium Member
Joined
·
308 Posts
Discussion Starter #19
IT WORKS! The 45% increase in rear beam stiffness is definitely noticeable. When you steer into a corner, you feel the car tuck in and respond more immediately than it does normally. This is exactly the effect I was hoping for. My finely-calibrated "butt dyno" tells me you could probably go to a 100% increase in stiffness and it'd still be fine. :) I drove on a variety of surfaces including several bumpy sections, some steep transitions that twist the rear beam heavily and didn't notice any increase in ride harshness.

I applied some primer to prevent rust and plan on driving a few days w/the bar installed to see if any issues arise. The mild steel used for this prototype was purchased at Home Depot and is probably far from the ideal material. I'll start shopping around for better materials including thicker wall square tubing for the muffler clamps + 4130 chromoly tubing for the cross bar & some stronger fasteners (grade 8 most likely) that'll hold up to be being twisted repeatedly. I may also look for an off-the-shelf U-bolt that'll work since fabricating my own took some time and the threads are totally unfinished meaning they'll start rusting badly.

One other note: the end of the muffler clamp touches the rubber brake line on both sides of the axle beam. To prevent chafing of the rubber tube, I added some spiral wire loom material which honestly is probably overkill but was super quick and easy to add so I figured it's better to be safe than sorry.
 

Attachments

1 - 20 of 36 Posts
Top