Jump to content




Holy Sphericals Batman!


  • You cannot reply to this topic
367 replies to this topic

#1
firstgencrx

  • This thing should fit!
  • PipPipPipPipPip
    • Group: 2013 Contributor
    • Location:opɐɹoloƆ 'suılloƆ ʇɹoℲ
This is a long post that's sharing what I'm working on, and my train of thought. Most of what I'm talking about here is pretty simplistic, so please don't get insulted. I just want people to be clear on what I'm trying to do. wink.gif

I managed to spend some time this week working on the spherical bearing kits I want to try and develop for the 1G/3G and 1G Integra's. These are the bearings that will replace the rubber/poly bushings at these locations....

Lower control arm rear joint here:


Torsion bar arm joint here:


and the rear axle trailing arms here:


My goal is to have these kits bolt in, but there may be some welding needed in some locations. Not 100% sure until I get it all worked out. These spherical bearings will effectively eliminate all flex at these points. This means if you are running these on the street, be prepared for some road vibration/noise transmitted up through the joint.

Every joint will have a grease fitting. The spherical bearings I've chosen to use are steel on steel, and will need to be greased for long life. In fact, if they are touched with a grease gun every few thousand miles, they should last for a very long time. The bearings also have some rubber seals on each side of the housing to help keep the dirt out. I just ordered the bearings this evening and I'm anxious to get my hands on them. They most likely will not ship out and start heading my way until Monday.

I ordered enough bearings to do two cars in all three locations. If they work out, I should be able to buy them in a large enough quantity to get a better price on them than what we see in the catalogs, and I always pass that savings on to you.

Now, I want to discuss a little more about the spherical bearing that will be in the location found in the second image. This is the joint at the front end of the torsion bar. This joint is going to be a little trickier than the others. My goal is to be able to service this joint without any special tools. You may still need a shop press to assemble it, and a typical bearing puller to get it apart, but no custom made tools like Honda uses to do the job. The spherical bearing I've chosen for this location is also very big. It's simply due to the fact that the stem on the back side of the torsion bar arm is a little over 1 3/8" in diameter.

The stem I'm talking about is this one:


The original rubber bushing that goes at this joint is basically an outer and inner length of tubing or sleeve that is vulcanized with rubber in between the two. The inner sleeve is pressed over the stem shown above, and then the outer sleeve is pressed into the end of the torsion bar tube which is part of the front sub-frame.

The outer part of the bushing is pressed into here:


Note: You can see the step machined inside the end of the torsion bar tube. This tube was bored like this at that the factory to make sure the size was consistent for a good repeatable bushing press fit.

The factory kept the torsion bar arm from falling off by relying on three things. The strength of the pressed-in inner and outer sleeves of the bushing, and the vulcanized rubber holding it all together. For safety, there is a snap ring out on the end of the torsion bar for insurance. I know some of you don't run the snap ring.

What is cool about Honda's decision to bore the end of the sub-frame tube, and turn down the stem on the back of the torsion bar arm for a good press fit for the bushings, is it allows me to take advantage of those same dimension. My goal is to have the spherical bearing press fit over the stem of the torsion arm, and press fit into a special sleeve that I make, that will then slide down the end of the torsion tube, and get bolted into place. HA! I think it'll work that way. tongue.gif I will also create some kind of device that will insure the assembly does not come apart. In practice sort of what Honda did with the snap ring, except with my device, it will not matter if you run with the snap ring.

The first part of the week has me anodizing some plate sets, but I know I'll have time to get started making the adapters and parts needed to make all of this work. I should be able to make most of it before I have the spherical bearings in hand, and just assemble everything when they get here. I'm hoping to have something to show by the end of next week or weekend. HA! I'm a dreamer. cool.gif

Have a great Halloween everyone!

David
Sharing with others what I do for myself!

My H22 CRX Project

Posted Image

#2
Jayscrx

  • Slowpoke
  • PipPip
    • Group: 2011 Contributing Member
    • Location:Clovis Ca
    • Drives: 06 Honda CRV , 08 Honda Fit Sport , 86 CRX Si
Hi David, I am very excited about this TB bearing kit that you are fabing up! I have the front suspension all apart to, replace all bushings, Teg radius arms, ball joints, 1 inch drop, new Tokico blue's, undercoat wheel well, and paint all suspension parts with VHT epoxy satin black with Delrin RA bushings. I could use this kit that you are working on very soon. Keep up the good work, very cool!

Jay V.

#3
Lymitliss

  • King Luke
  • PipPipPipPipPip
    • Group: 2012 Contributor
    • Location:Seattle, Wa
    • Drives: 1986 CRX Si B18C1, 1973 240Z V8 Conversion
    Garage View Garage
Great idea, I'm very excited to see how it turns out.
Build Thread- Here


#4
EPcivic

  • 2008 & 2009 EP National Champion
  • PipPipPipPip
    • Group: 2010 Contributing Member
    • Location:Findlay, Ohio
I'm very interested in the radius rod bushing, as I think it has more of an effect on handling then the others. One thing to consider - I've always thought that you couldn't replace both front bushings with spherical joints because they are rotating about different centers, yet they are bolted solid at the knuckle end. Maybe the torsion arm is compliant enough in fore-aft bending that it wouldn't matter. I'd love to hear from someone who has done sphericals in both locations, and see how it worked out. I'm not saying that it won't work, but I've never looked into doing it because I can't visualize where the torque arm and radius rod will deform to make up for the removed radial compliance in the bushing.

-Chris

#5
DEIVIONCRX

  • Demon Of All Trades
  • PipPipPipPipPipPip
    • Group: 2013 Contributor
    • Location:Oklahoma
    • Drives: 87 Civic Si, 88 Civic DX, 90 CRX DX, 80 Trans Am, 10 Fit Sport
    • Image Gallery
    Garage View Garage
If you get this to work, you will be the most awesome guy in the planet David.

Give me those 3 bushings in spherical and the rest i can deal with being, rubber/poly/delrin. Which now that i think about it, between your panhard, your camber plates, and a HF axle, leaves only the upper rear shock mounts, lol.

EP, the sphericals should allow both arms to rotate and move in any direction they want, without binding. Its just a simple ball/socket joint as you probably well know, it should be just fine, living with a spiral type pattern as it rotates.
Posted Image

#6
CrxTom

  • Slowpoke
  • PipPip
    • Group: 2010 Contributing Member
    • Location:Midcoast Maine, USA
    • Drives: 1985 CRX
Awesome idea. Will be watching this thread closely (drool)

#7
EPcivic

  • 2008 & 2009 EP National Champion
  • PipPipPipPip
    • Group: 2010 Contributing Member
    • Location:Findlay, Ohio
QUOTE (DEIVIONCRX @ Oct 31 2009, 02:25 PM) <{POST_SNAPBACK}>
EP, the sphericals should allow both arms to rotate and move in any direction they want, without binding. Its just a simple ball/socket joint as you probably well know, it should be just fine, living with a spiral type pattern as it rotates.



Take a closer look at the system. A sphericial joint allows rotation axially, and rotation about the ball of the joint, but no radial movement. The two suspension arms a bolted together at the steering knuckle end. The torque arm rotates about the torsion bar. The radius rod rotates about a point that is a lesser distance from the steering knuckle, so if you fix the torque arm to pure rotation, then the radius arm needs to travel in a slight arc while it rotates. If you fix both in pure rotation, then the combination of arms must deform to allow them to rotate. I'm just not sure how the arms would deform. The torque arm is very stiff up/down and in/out, but very soft fore/aft. The radius rod is stiff fore/aft, but very soft in/out, and up/down.

A lot of cars use this same sort of lower control arm configuration - it allows the Engineers to tune the bushing compliance more easily. Often, on race cars, you will see a radius rod with spherical joints at both ends, which fixes the binding / bending issue I am concerned about.

-Chris

#8
firstgencrx

  • This thing should fit!
  • PipPipPipPipPip
    • Group: 2013 Contributor
    • Location:opɐɹoloƆ 'suılloƆ ʇɹoℲ
Maybe this will help everyone else see what you guys are talking about.

Like EPcivic says, the two arms rotate on different radii. The torsion bar arm being the longer one, and the rear joint on the lower control arm being the shorter one. Here is a simplified view of what I mean. The red lines represent the two different axis of rotation:



Because we have rubber at these two joints, and the torsion arm can twist, the system created by these two arms bolted together has a combined axis of rotation that looks and behaves something like this:



This is not the exact behavior, but I would bet it's pretty damn close.

The lower control arm is built like a ship. It's not flexing anywhere except under the very heaviest loading. The torsion bar arm can only twist because it's so thin. It can not change in length.

I'm really trying hard to use spherical bearings at both locations simply because they will offer the least (zero really) amount of binding. This should make the whole system stronger, and smoother acting. This setup will also remove the variable spring rate that is added to the front caused by the torsion arm rubber bushing. It's much more than you would think. When I was doing my bump steer data collection, the force to fully compress the front without a torsion bar, with only the spring resistance of the torsion bar bushing was considerable. I'm going to guess it's something greater than 300 pounds out on the ball joint end of the LCA. Does this effect the handling?

HA! It just occurred to me, the combination of running front spherical bearings and a bump steer correction kit might change enough of the handling behavior that some of the old tried and true tuning tips might need a little tweaking. tongue.gif

David
Sharing with others what I do for myself!

My H22 CRX Project

Posted Image

#9
DEIVIONCRX

  • Demon Of All Trades
  • PipPipPipPipPipPip
    • Group: 2013 Contributor
    • Location:Oklahoma
    • Drives: 87 Civic Si, 88 Civic DX, 90 CRX DX, 80 Trans Am, 10 Fit Sport
    • Image Gallery
    Garage View Garage
The sphericals your talking about are just like a rod end right, put the torsion bar arm through the center ball, so it can rotate 360*, and it can pivot whatever like 37*, so as the suspension compress's it can rorate and pivot as it wants.

I just dont get where the problem is cuz the spherical joint allows it to rotate on any axis it wants, the only thing stoping it is the torsion bar.

And then by adding the spherical on the Radius arm, that bar isnt holding it back from rotating on the axis it wants.
Posted Image

#10
firstgencrx

  • This thing should fit!
  • PipPipPipPipPip
    • Group: 2013 Contributor
    • Location:opɐɹoloƆ 'suılloƆ ʇɹoℲ
QUOTE (DEIVIONCRX @ Oct 31 2009, 11:58 AM) <{POST_SNAPBACK}>
The sphericals your talking about are just like a rod end right, put the torsion bar arm through the center ball, so it can rotate 360*, and it can pivot whatever like 37*, so as the suspension compress's it can rorate and pivot as it wants.

I just dont get where the problem is cuz the spherical joint allows it to rotate on any axis it wants, the only thing stoping it is the torsion bar.

And then by adding the spherical on the Radius arm, that bar isnt holding it back from rotating on the axis it wants.


Correct. It's just like the end of a spherical rod end, or Heim as some call it.

The joint at the end of the torsion bar will actually rotate at the same axis as the torsion bar. This is simply because to do different, it would need to bend the torsion bar, and I doubt that's going to happen. There is a little play in the spline, but with the weight of the car loading the system, I just cant see the end of the torsion bar arm tipping much there. I could technically use just a ball bearing at that location. The problem is the loading. A ball bearing that would fit inside there is pretty wimpy, where the spherical bearing has a mega load rating. A needle bearing might be an alternative, or even something as simple as a bronze bushing. In the past, I didn't think something like a bronze bushing would work at the end of the torsion bar, but after spending more time with this, I think it would.

The problem I see with using a needle bearings or a bronze bushing instead of a spherical bearing is keeping the torsion bar arm from coming out of there. A system of retention would need to be developed that is more than just a safety device. It would need to incorporate thrust retention as well. I'm hoping the spherical bearing results in the fewest number of parts for the kit, which translates into the lowest cost to all.

The greatest benefit of this work is to remove the flex from the two joints caused by the rubber/poly bushings.

David
Sharing with others what I do for myself!

My H22 CRX Project

Posted Image

#11
CrxTom

  • Slowpoke
  • PipPip
    • Group: 2010 Contributing Member
    • Location:Midcoast Maine, USA
    • Drives: 1985 CRX
Perhaps if this is successful (as I think it will be), and if you wind up making these for sale, maybe people could opt-out of this specific piece if it made them uncomfortable. Just an idea wink.gif

If the price is right, I'll be all over this biggrin.gif

#12
firstgencrx

  • This thing should fit!
  • PipPipPipPipPip
    • Group: 2013 Contributor
    • Location:opɐɹoloƆ 'suılloƆ ʇɹoℲ
I plan on letting people pick and choose the joints they want to replace with spherical bearings.
Sharing with others what I do for myself!

My H22 CRX Project

Posted Image

#13
kakabox

  • Turd Polisher
  • PipPipPipPipPip
    • Group: 2012 Contributor
    • Location:Seattle, WA 98115
    • Drives: from point 'A' to point 'B' daily.
    • Image Gallery
    Garage View Garage
QUOTE (firstgencrx @ Oct 31 2009, 10:48 AM) <{POST_SNAPBACK}>
Because we have rubber at these two joints, and the torsion arm can twist, the system created by these two arms bolted together has a combined axis of rotation that looks and behaves something like this:



This is not the exact behavior, but I would bet it's pretty damn close.

The lower control arm is built like a ship. It's not flexing anywhere except under the very heaviest loading. The torsion bar arm can only twist because it's so thin. It can not change in length.

I'm really trying hard to use spherical bearings at both locations simply because they will offer the least (zero really) amount of binding. This should make the whole system stronger, and smoother acting. This setup will also remove the variable spring rate that is added to the front caused by the torsion arm rubber bushing. It's much more than you would think. When I was doing my bump steer data collection, the force to fully compress the front without a torsion bar, with only the spring resistance of the torsion bar bushing was considerable. I'm going to guess it's something greater than 300 pounds out on the ball joint end of the LCA. Does this effect the handling?

HA! It just occurred to me, the combination of running front spherical bearings and a bump steer correction kit might change enough of the handling behavior that some of the old tried and true tuning tips might need a little tweaking. tongue.gif

David

David,
I agree w/the torsion bar arm being able to bend/twist due to it's geometry.

I think what EPcivic is saying is that w/a spherical bearing in the tbar arm, that 'blade' will see a more bending/twist because of the reduction in radial translation that the OE rubber bushings was providing. The increase in the bending/twisting load will put more radial load on the spherical bearings in the tbar and radial arm and in the two bolt joint connecting the two. Something to think about.

Look at your second pic above. If the axis of rotation is the red line, then the tbar will have have to cock itself (move) within the confines of it's splines/tube to rotate the spherical bushing to achieve this axis of rotation. So will the end of the radial arm. However, w/out any radial translation (compliance) available, how does this occur? I don't think there's enough room within the free-play of the splines or the tube to allow this. Hence, the radial load will go up compared to the OE system, bending/twisting the tbar arm more than OE.

Just my thoughts for now...I'll continue to mull it over.

#14
firstgencrx

  • This thing should fit!
  • PipPipPipPipPip
    • Group: 2013 Contributor
    • Location:opɐɹoloƆ 'suılloƆ ʇɹoℲ
Your thoughts are good! I am seeing what you are saying, and I agree, there is more twisting needed by the torsion bar arm.

The system like I said above is more complicated than my simple illustration. I'm trying to find a way to draw what really happens, but I'm at a loss. I suspect there are actually two axis of rotation. One axis of rotation is down the torsion bar, and the second is at the rear joint of the lower control arm, pointing at the front torsion bar joint. Here is a picture I made that might resemble what I'm trying to say:




I feel the axis of rotation right at the spline interface between the torsion bar arm where it slips onto the end of the torsion bar is parallel to the torsion bar. The reason I explain in post #10 is simply because for the end of the torsion bar arm to articulate any at that spline interface would mean the end of the torsion bar would have to bend. I don't think the torsion bar end is bending. We know that there is play in the spline interface, but when you apply the extreme torque to the spline from loading the car, my guts tell me the end of the torsion bar arm is not able to rock on the end of the torsion bar.

Definitely the torsion bar arm is twisting to make up for the two combined axis of rotation. How much more than with a rubber bushing at both joints? I'm guessing not a lot. With the rubber bushings, the torsion bar arm still needs to twist and bend to make the system work. My guts tell me with the spherical bearings, the arm will not see much more of that behavior.

Ideally, it would be cool to completely decouple the splined end of the torsion bar arm from the end of the torsion bar like they do with most race cars. But that would require major changes to the front sub-frame.

David

NOTE: It's easy enough to measure how much more or less the torsion bar arm will twist and bend with and without the spherical bearings. I'll just measure it. wink.gif
Sharing with others what I do for myself!

My H22 CRX Project

Posted Image

#15
EPcivic

  • 2008 & 2009 EP National Champion
  • PipPipPipPip
    • Group: 2010 Contributing Member
    • Location:Findlay, Ohio
I agree that we aren't talking about a lot of movement here, but I still think that something has to 'give' if you eliminate all of the radial play in both bushings without adding another degree of freedom, like a spherical bearing between the radius arm and torque arm at the knuckle.

My guess is that if you were to fab up a quick bushing to hold the torque arm to purely rotational motion, you could check the resulting path at the other bushing location pretty easily. Regardless of the type of bearing you put at that location, I think it is only going to have rotational freedom, because the spline fit and torsion bar aren't going to allow for much other movement.

My gut tells me that the radius rod travels in a slight arc up & down. At some point in the travel, I would bet it is concentric with the subframe opening. As you move away from that point, you will get a spring rate due to something in the system flexing. This may be negligable, or it may completely lock up - I really don't know. I would definately be interested in seeing the results of the measurement. I would also be interested in knowing where the 'zero' or concentric point is, and if the normal operating range is going to experience both positive and negative spring forces due to the binding.

-Chris