Bible Linked Suspensions Bible

[Weasel]

nuance is what I was going for.

 

[Treefrog]

Taking a break from 3d for a bit to go back to exploring 2d suspension. Managed to simplify the geometry solving code some by solving as though it exists in 3d., though I have not timed the difference. But those changes are for a different time and thread.

Skip to after the line for tech stuff.

The primary focus so far has been on analyzing forces instead of geometry. So far I am only looking at it from a static and instantaneous 1 g acceleration perspective. Over the next weeks/months, I hope to have some notes on steady state acceleration, traction limited instantaneous and steady state acceleration, and add in uphill and down hill slopes to all of it.

One of the current limitations of my forces modeling is that I only have axle mounted shocks set up so far, but it should not have to much of an effect, yet.

---

I do not have any images ready to go with this post, but should have something to show later in the week.
Looking briefly at a rig at rest with no brakes. In particular, the load distribution on the front and rear. Historically, we have said that the weight bias is derived from the CG location as a percent of wheelbase. However, this is not 100% accurate. In reality, the spring causes a moment on the axle housing that results in loading of the links. From the one or two cases I have looked at so far, the difference between weight bias and force bias is a few lbs at most.

Looking at the instantaneous acceleration, or before the suspension has had time to move. I'll have an image for this in a few days. I took a look at the overall reaction in the vertical and pitch for a 1g acceleration as the drive bias changes from 100% rear to 100% front. The interesting part was, for the suspension I analyzed, the direction of the change in vertical force changed as the bias changed. I do not think it had 50/50 as the change over point. The direction of the moment never changed.

I am curious to see the analysis for steady state acceleration because the instantaneous force shift did not match the results from the load transfer equation.

 

[Treefrog]

It seems I may have been wrong about the CG location and force at both ends, so I crossed it out. I guess I was remembering the results during troubleshooting.

Now onto some pictures. I am fairly confident in the results, but not 100%. I probably should draw a FBD. All of these plots are for an instantaneous 1 g acceleration varying the % of force applied by each end.

Some basic setup info:

total weight: 3446 lbs​

front weight at rest: 1835.7 lbs​

rear weight at rest: 1610.6 lbs​

no portals​

Calculated 100% bias to relevant end anti values:

front:76.3%​

rear: 88.5%​

Up first is a side view of the suspension used. +X is forward.

The moment on the chassis:

The forces on the links, in 2d. I did not correct them into the actual 3d force on the link.

But those results are not that surprising or interesting. Nose goes up and lower links have larger magnitudes. The vertical forces are much more interesting. First the vertical force on the chassis.

Forces on the ground, acceleration, static, and expected load transfer.

And last up, the difference in force to the ground from rest.

 

[drozg]

Can we talk about the reverse 4 link configuration (straight lowers and uppers converging at the chassis), that more and more leading Ultra4 builders started using in their latest cars - Triton, Horschel…

So, I am wondering what is the advantage of that configuration?

Most of them have mentioned that the car feels more stable, but why?

I am sure it has something to do with the roll center, as I have noticed that with that configuration the roll center curve through the travel is reversed (roll center gets higher with down travel) compared to traditional setup where roll center gets higher with up travel. Does that make the car more stable?

I have always thought that for stable car you want as little roll center height change through travel as possible, and as flat vehicle roll axis as possible, which I can both achieve with traditional triangulated setup… Or am I wrong and is the reversed roll center curve superior for stability?

 

[Treefrog]

Can we talk about the reverse 4 link configuration (straight lowers and uppers converging at the chassis), that more and more leading Ultra4 builders started using in their latest cars - Triton, Horschel…

So, I am wondering what is the advantage of that configuration?

Most of them have mentioned that the car feels more stable, but why?

I am sure it has something to do with the roll center, as I have noticed that with that configuration the roll center curve through the travel is reversed (roll center gets higher with down travel) compared to traditional setup where roll center gets higher with up travel. Does that make the car more stable?

I have always thought that for stable car you want as little roll center height change through travel as possible, and as flat vehicle roll axis as possible, which I can both achieve with traditional triangulated setup… Or am I wrong and is the reversed roll center curve superior for stability?

I am more than happy to look into it. But I do not think I will have time to until sometime next week.

Have any of the drivers said in when it is more stable? Off camber, at speed, sliding in a corner?

I think it has more to do with the roll slope than with the roll center. From some reading I was doing over the weekend, the roll center is not a point of rotation, but a point at which an applied force does not result in roll.

 

[Treefrog]

As a follow up to the force plots and 2d stuff. Looking at acceleration vs drive/brake bias by adjusting the used coefficient of friction from -1 to 1. Negative coefficient of friction is braking as far as the code is concerned. A straight dotted line between the two end points is present for clarity. It is interesting that the results are not a straight line.

And pushing the friction to -3 to 3 out of curiosity.

edit: The equations and methods I am using to generate these plots should be adaptable to the publicly available 4 link calc when I get around to expanding its capabilities as some point.

 

[drozg]

I am more than happy to look into it. But I do not think I will have time to until sometime next week.

Have any of the drivers said in when it is more stable? Off camber, at speed, sliding in a corner?

I think it has more to do with the roll slope than with the roll center. From some reading I was doing over the weekend, the roll center is not a point of rotation, but a point at which an applied force does not result in roll.

Awesome, thank you!

Unfortunately they didnt want to be too specific with the informations.

One of them said that ‘’the roll center is not all over the place during travel, like with traditional setup, which gives you more stable roll axis’’

However like I said, with double triangulated you can get almost constant roll center during travel (only moves 2-3 inches over 18 inches of travel, based on the calculator of course), when reversed move more, but in opposite direction than traditional setup.

So now I am wondering, if this is even better or not? I suspect something should be about it that works better, if they have tried it all and liked the reversed system the most.

 

[Motiracer38]

A lot of short course trucks have the uppers reversed and have been for years so there must be a reason. It would be nice to nail down the characteristic it favors. A few trophy trucks tried it as well but it never took off, not sure whether due to handling, packaging, or reliability.

 

[Treefrog]

Unsurprisingly, I've got a bit of a to do list of things to figure out, verify, or prove/dis-prove when it comes to 4 links. And one that has been at the top of the list has been is: How does antisquat relate to chassis forces and moments? In particular, one of the big questions has been if 100% antisquat actually results in increased traction and no squatting of the rear suspension. I am now able to say that in general, 100% antisquat does NOT mean no squatting. In general, it does appear to result in increase rear traction. It seems that antisquat is only barely related to pitching while the IC is significantly more important.

One of the things that I have never been quite sure of is what CG height to use for antis. To my knowledge, the calculator has always used the combined CG of the sprung mass and the un-sprung mass of the opposite end. But, in the event that is not right, I also looked at using the sprung CG and the vehicle CG.

For all of the plots, the only data point that matters is the 100% rear drive. I did not set the front to 100% anti. Moments are in in*lbs. I did not catch it before while making the graphs.
First is sprung and opposite un-sprung CG.

Second is vehicle CG.

Fourth is sprung CG.

All three show a front-up moment and increased contact force.

So where does the idea that it means no vehicle pitching come from? My guess is that this is one of the many cases of textbook suspension design not being focused on tall, solid axle vehicles. It is likely that a short, IFS/IRS vehicle with close to 100% antisquat will not show any noticeable pitch.

 

[Treefrog]

Looking into uppers that are more narrow at the frame. The results are not quite what I expected and are interesting from a "more stable" view.

I want to start off by saying that my view on roll slope and roll center (and antis) has been gradually moving towards the "not real" camp. But that is a rabbit hole to go down. From playing around with the calculator, it is possible to get a suspension to have all sorts of trends. For example, close together uppers at the axle have very little roll center movement with travel. The one of the big problems with roll center and the like is that they are rough predictions.

Finding info on having the uppers narrow at the frame, was very difficult. There may be a few threads from the early 2000s on PBB, but there are dead links and missing pictures involved. Another thing I looked into as part of it is the Satchell link that gt1guy mentioned in post 32. Satchel links are straight uppers and traditional triangulated lowers. At the time the discussion was on anti while turning.

The first thing I did was try to find examples of these cars as reference. The examples from above regarding uppers that narrow at the frame are some trophy trucks, short course trucks, Triton, and Horschel. An example of a Satchel link are the UFO cars, though their uppers are slightly closer at the frame.

What is interesting is that I did not find any examples of them on front solid axle rigs, but I also did not search to hard. I think there is a reason for this. But more on that later.

When it comes to stability, there are a few things that come to mind. Is the rear end roll steering less? Is the outside tire gripping less when cornering promoting drifting? Is it resulting in less body roll? Is it less bucking of a corner. Among others of decreasing likely hood.

The two values I chose to look at in this study are roll steer and the side to side movement of the center of the axle.

The side view of the suspension is the same in all cases. It is a generic U4 setup. Decently long with flat lowers and uppers that are a bit lower at the frame. I set the solved condition of the suspension to vertical travel of the center of the wheels.

Six setups were compared. From a top view the normal setup is a dual triangulated setup with the uppers having most of the angle. Inverted switched the Y values of the upper link's endpoints. Wide is straight uppers at the wider of the normal's upper Y values. Narrow is straight uppers at the narrower of the normal's upper Y values. Fifth is wide with more angle in the lowers. And last is switching the Y values between normal's upper and lower links.

So onto some numbers and images. Roll centers and slopes are provided. Positive slope is oversteer. Roll steer first.

And center shift.

The color scales are the same for plots of the same info.

The trend for these and other suspensions I have played with with lower links near the axle centerline:

• Wide at axle for upper and/or lower result tend to have lower roll steer
• Straight uppers seems to have lower center shift

The reason I think that they are saying more stability is mostly related to the roll steer.

As for satchel links, I think there is some predictability with lower center shift. With solid front and rear, the both axles move side to side with body roll, thus axle center to axle center line stays straight. But IFS does not really have the front center shift. I believe Wayne referenced this on his YouTube channel Wayne's Way when talking about lifted Ford Broncos. He said that the panhard rear moves the rear end around and results in some odd handling while Jeeps do not have the same handling quirkiness despite having the same panhard issue because the front also moves around.

It does seem to be possible to keep both roll steer and center shift low with some design iteration.

The other thing to keep an eye on is where the roll steer is in travel. With dual rate springs, you will have more down travel than up travel for a given body roll. It would be cool to put some sensors and data logging on a U4 car to see where in travel it spends most of its time.

I think part of it has to do with packaging. It is much easier to get wide spacing of the uppers on the axle if they are narrow at the chassis when running trailing arms.

 

[ditcherman]

One of the things that I have never been quite sure of is what CG height to use for antis. To my knowledge, the calculator has always used the combined CG of the sprung mass and the un-sprung mass of the opposite end. But, in the event that is not right, I also looked at using the sprung CG and the vehicle CG.

......clipped

So where does the idea that it means no vehicle pitching come from? My guess is that this is one of the many cases of textbook suspension design not being focused on tall, solid axle vehicles. It is likely that a short, IFS/IRS vehicle with close to 100% antisquat will not show any noticeable pitch.

I want to start off by saying that my view on roll slope and roll center (and antis) has been gradually moving towards the "not real" camp. But that is a rabbit hole to go down. From playing around with the calculator, it is possible to get a suspension to have all sorts of trends. For example, close together uppers at the axle have very little roll center movement with travel. The one of the big problems with roll center and the like is that they are rough predictions.

........clipped

First of all let me say that as a total newb to all this I find it fascinating, even if I can barely understand it. I also find I'm understanding it better the second (at least) time through, since I've built (and unbuilt, and built, and unbuilt, and built) a suspension. I'm getting really close and I can't wait to drive it and post up in the numbers thread how great it turned out, since it feels like I've been chasing numbers for months.

Which leads me to my question and why I quoted portions of those two posts, as you're exploring things that you think you knew but aren't sure about now....The calculator builds good trucks, right? (oh please tell me the calculator builds good trucks!!) Seriously though, from my 10,000 foot view, you've developed and tweaked the calculator to build good suspensions. If someone builds a good suspension without it they either knew what to do, or got lucky. So if the calculator builds good suspensions, is there a way to compare bad suspensions to good ones, to prove what you are asking here, if anti's are even a real thing, if the CG's used are right?

I believe the challenge in this would be the fact that someone who built a suspension wrong doesn't care enough to use the calculator let alone share the bad design, but maybe it could happen.

Just curious, as I learn about all this and watch the evolution of this subject over the years of this thread.

Your work is appreciated and I'm glad to be here, now, at this moment so in my build I am not as fixated on the anti's as I would have been a few years ago.

 

[Treefrog]

First of all let me say that as a total newb to all this I find it fascinating, even if I can barely understand it. I also find I'm understanding it better the second (at least) time through, since I've built (and unbuilt, and built, and unbuilt, and built) a suspension. I'm getting really close and I can't wait to drive it and post up in the numbers thread how great it turned out, since it feels like I've been chasing numbers for months.

Cool. Closing the loop on how it performs vs. what the numbers say is the most important part right now. The more feedback we can get the better understanding we have for the next guy that comes along asking about suspension.

Which leads me to my question and why I quoted portions of those two posts, as you're exploring things that you think you knew but aren't sure about now....The calculator builds good trucks, right? (oh please tell me the calculator builds good trucks!!)

Nope . The calculator can help design a bad suspension just as easily as it does a good one. All it is is a tool, like a welder or a hammer.

Seriously though, from my 10,000 foot view, you've developed and tweaked the calculator to build good suspensions.

Not quite. I have added stuff to the calculator that I think is helpful in designing a suspension.

If someone builds a good suspension without it they either knew what to do, or got lucky.

I disagree. A good suspension is easy. The general guidelines in five minute read magazine articles that have been floating around for close to two decades now have held up to the test of time pretty well. Designing a great suspension that knowledge, experience, or luck.

So if the calculator builds good suspensions, is there a way to compare bad suspensions to good ones,

User feedback. More data points. 1 guy saying here is my numbers, this is how it handles provides some insight and can make you think. 100 guys providing the same stuff allows for conclusions and understanding.

to prove what you are asking here, if anti's are even a real thing,

Antis exist. The trend is high antisquat will resist squatting more. The question is how accurate are they at predicting how much squat. And if 100% is actually no movement. What I can do is generate a dataset that says X anti with a IC this far from the axle has Y moment and Z vertical force. I can then talk about what the data shows. What I think I will find is that it will depend on if the extended line of the links pass above or below the chassis CG. Which in a sports car probably reflects the anti values fairly close.

if the CG's used are right?

I think this one is just a do more reading sort of solution. I am guessing the answer is in some textbook or SAE document. I have only briefly tried tracking down the origin of anti-squat. And just now checking Milliken and Milliken's book, it appears that their derivation uses the entire vehicle's CG.

I believe the challenge in this would be the fact that someone who built a suspension wrong doesn't care enough to use the calculator let alone share the bad design, but maybe it could happen.

Plenty of suspensions that we would caution against today were built with the calculator. Some were redone, and some are still out there today.

Just curious, as I learn about all this and watch the evolution of this subject over the years of this thread.

Your work is appreciated and I'm glad to be here, now, at this moment so in my build I am not as fixated on the anti's as I would have been a few years ago.

Thanks. I am also curious where it will go in the future.

 

[Treefrog]

I am bringing an old post back up from the dead Treefrog This isnt for me but for a friend building a 4500 car. He has his lowers triangulated in a front 3 link and is concerned about bumpsteer at speed. Does triangulating the lowers cause more bumpsteer even if the trackbar and drag link are equal and flat?

When does inboarding become too much?

I’ll chime as I’m the one with the front 3 link for a 4500 car.

Numbers are still adjustable to an extent

6.5 inches vertical separation at frame
8.5 at axle

Center of heim to center of heim lower links at axle is roughly 39.5

At the frame is 26.5

I don’t have the panhard or draglink done yet. I have room to move the frame side lower links out maybe an inch or two each way if I redo the subframe some. I can lower the frame side upper link as well some too.

How much will having the frame side lowers triangulated affect bump steer while articulated and while at speed?

Thanks

Hope you do not mind me moving this to this thread. It seems that this is more theoretical and I try to keep that stuff here. If we start to look at actual coordinates in space we should move back. The numbers provided do not do a great job of describing where stuff is in space.

I think this may need to be split into a discussion on minimizing roll steering (and travel steering) of an axle with a panhard and minimizing bump steer.

As for how does lower triangulation effect bump steer when a panhard is in use, it is not great. I believe that from a geometry aspect, narrow at he frame lowers have the greatest potential to cause roll steer. The only reason it may not show up is if side to side movement of the center of the axle is kept to a minimum.

To picture this, imagine a 1 link and panhard suspension. The axle rotates about a point on the chassis. As it moves side to side from the panhard, it will always point to the rotation point.

As the lower links move from parallel to close at the frame, the suspension looks more and more like the one link.

The interesting thing is that the roll steer will be very heavily biased to one side. It would take some number crunching, but it may be possible to use a panhard that is shorter than the drag link and positioned correctly to counter this. I do not think that the position is parallel to the drag link. Some caution would need to be taken at the limits of steering angle.

One thing that could be looked into is having the axle end of the panhard forward of the frame mount. As the axle travels through its arc, the panhard will effectively be adding length, lowering side to side movement. I think the calculator shows this but I am not certain.

---

I know it was not the targeted topic, but I think I should comment on the drag link and panhard relationship. The equal and flat is good guidance but is not absolute. For example, unequal up down travel should typically point towards a not flat panhard. Additionally, the equal and parallel is only truly correct when the upper and lowers are equal length and parallel.

 

[Malburg114]

Hope you do not mind me moving this to this thread. It seems that this is more theoretical and I try to keep that stuff here. If we start to look at actual coordinates in space we should move back. The numbers provided do not do a great job of describing where stuff is in space.

I think this may need to be split into a discussion on minimizing roll steering (and travel steering) of an axle with a panhard and minimizing bump steer.

As for how does lower triangulation effect bump steer when a panhard is in use, it is not great. I believe that from a geometry aspect, narrow at he frame lowers have the greatest potential to cause roll steer. The only reason it may not show up is if side to side movement of the center of the axle is kept to a minimum.

To picture this, imagine a 1 link and panhard suspension. The axle rotates about a point on the chassis. As it moves side to side from the panhard, it will always point to the rotation point.

As the lower links move from parallel to close at the frame, the suspension looks more and more like the one link.

The interesting thing is that the roll steer will be very heavily biased to one side. It would take some number crunching, but it may be possible to use a panhard that is shorter than the drag link and positioned correctly to counter this. I do not think that the position is parallel to the drag link. Some caution would need to be taken at the limits of steering angle.

One thing that could be looked into is having the axle end of the panhard forward of the frame mount. As the axle travels through its arc, the panhard will effectively be adding length, lowering side to side movement. I think the calculator shows this but I am not certain.

---

I know it was not the targeted topic, but I think I should comment on the drag link and panhard relationship. The equal and flat is good guidance but is not absolute. For example, unequal up down travel should typically point towards a not flat panhard. Additionally, the equal and parallel is only truly correct when the upper and lowers are equal length and parallel.

Thanks for the reply.

So to make sure I understand correctly, to minimize the bump/roll steer since the lowers are more triangulated, cycle the suspension with a slightly shorter panhard and moving the axle end of the panhard forward of the frame side one or I guess even the frame side one back. How much would you shorten the panhard and how much offset? Or is it a more trial and error of trying different combinations?

I’ll cycle it like my last build with it all equal to the drag link and see if I get any actual bump steer before messing with it as I was sure the other one would have bumpsteer with a shorter panhard but there was next to none (set a camera looking at the steering wheel and the wheel barely moved a 1/16 throughout the travel)


My general understanding of your comment is minimize the side to side swing of the axle throughout the travel and the less bump steer it’ll have. Correct?

 

[Treefrog]

Thanks for the reply.

So to make sure I understand correctly, to minimize the bump/roll steer since the lowers are more triangulated, cycle the suspension with a slightly shorter panhard and moving the axle end of the panhard forward of the frame side one or I guess even the frame side one back. How much would you shorten the panhard and how much offset? Or is it a more trial and error of trying different combinations?

A shorter panhard than drag link MAY help with bump steer. Moving one end front or back MAY help with roll steer and side to side movement.

Without running numbers in 3d or doing trial and error in CAD, trial and error cycling the suspension is your best bet. The 4 link calc can get close, but it is not able to solve in 3d.

I’ll cycle it like my last build with it all equal to the drag link and see if I get any actual bump steer before messing with it as I was sure the other one would have bumpsteer with a shorter panhard but there was next to none (set a camera looking at the steering wheel and the wheel barely moved a 1/16 throughout the travel)


My general understanding of your comment is minimize the side to side swing of the axle throughout the travel and the less bump steer it’ll have. Correct?

Minimizing side to side movement keeps the axle square to the chassis during pure up down travel. You then need to locate the drag link to minimize bump and roll steer.


I want to create some plots showing the axle squareness, but the CPU cooler in my computer has decided to no longer cool. I should be able to get plots Monday or Tuesday night.

 

[Treefrog]

Here is some plots. The general setup was flat lowers, slightly angled in at axle upper on frame side of the panhard. Upper was slightly lower at the frame. Actual results seem to be highly dependent on all of the links more than they would be for a triangulated 4 link.

Wide frame lowers with no forward back offset of the panhard:

Wide frame lowers with slightly rear at the frame panhard:

Narrow frame lowers and no forward offset of the panhard:

Narrow frame lowers with slightly rear at the frame panhard:

 

[Treefrog]

A question crossed my mind this morning. On a front and rear solid axle rig, would it be better to tolerate and/or prefer side to side shifting of the axle if it shifts in the direction of the body lean? As in, the center of the axle moves downhill relative to the chassis and by extension back under the CG.

 

[gt1guy]

A question crossed my mind this morning. On a front and rear solid axle rig, would it be better to tolerate and/or prefer side to side shifting of the axle if it shifts in the direction of the body lean? As in, the center of the axle moves downhill relative to the chassis and by extension back under the CG.

I'd say no. And definitely not on a long travel offroad rig. One reason, the front and rear may not move exactly the same at the same time. That would cause some sketchy handling quirks.

Would it not also happen during a corner?

So if the axle moved laterally with the intention of reducing some of the roll, as soon as the axle stops moving laterally, wouldn't all the roll return. Depending how sudden the stop in lateral movement is, there could be a dump of roll induced into the chassis.

At 2mph it might work

 

[Treefrog]

I'd say no. And definitely not on a long travel offroad rig. One reason, the front and rear may not move exactly the same at the same time. That would cause some sketchy handling quirks.

Would it not also happen during a corner?

So if the axle moved laterally with the intention of reducing some of the roll, as soon as the axle stops moving laterally, wouldn't all the roll return. Depending how sudden the stop in lateral movement is, there could be a dump of roll induced into the chassis.

At 2mph it might work

Corners are what happen when you flatten a side slope.

The intention would be increasing the margin for roll over, not reducing the amount of roll.

I may have worded it poorly. The natural movement of a 4 link suspension involves the center of the axle moving away from the chassis's center plane.

I don't view it as shifting the axles to influence roll. Instead, it is roll changing the relative positions of the chassis and axles.

 

[Weasel]

Corners are what happen when you flatten a side slope.

The intention would be increasing the margin for roll over, not reducing the amount of roll.

I may have worded it poorly. The natural movement of a 4 link suspension involves the center of the axle moving away from the chassis's center plane.

I don't view it as shifting the axles to influence roll. Instead, it is roll changing the relative positions of the chassis and axles.

Unsure that the margin for roll over has much to do with the geometry or chassis movement in general, there is a rate to the roll over point that you can mess with. May not be following the question/thought.

Antis exist. The trend is high antisquat will resist squatting more. The question is how accurate are they at predicting how much squat. And if 100% is actually no movement. What I can do is generate a dataset that says X anti with a IC this far from the axle has Y moment and Z vertical force. I can then talk about what the data shows. What I think I will find is that it will depend on if the extended line of the links pass above or below the chassis CG. Which in a sports car probably reflects the anti values fairly close.

They do but what I don't quite follow is that Anit's only come in during acceleration. Any steady state situation you don't have them which is the majority of time a vehicle is in motion (atleast for crawling).

 

[Treefrog]

Unsure that the margin for roll over has much to do with the geometry or chassis movement in general, there is a rate to the roll over point that you can mess with. May not be following the question/thought.

What I am thinking is that you have two vehicles, A and B. A does not have any side to side movement of the center of the axles. Whereas for B, the center of the axles moves 3 inches from the chassis centerline towards the compressed side. Both have the same chassis and axle CGs. Vehicle A will roll over with less sideways Gs (from the axles' reference frame) than vehicle B.

They do but what I don't quite follow is that Anit's only come in during acceleration. Any steady state situation you don't have them which is the majority of time a vehicle is in motion (atleast for crawling).

Squatting, lifting, and torque roll are the result of drive forces not acceleration. Acceleration usually produces the highest force. There is force in the form of drag (air and driving surface) during steady state. Steady state uphill driving is another force example.

 

[gt1guy]

How are we getting the axles to move laterally 3" to either side? Do both the front and rear axles move the same amount at the same time?

 

[Treefrog]

How are we getting the axles to move laterally 3" to either side?

The suspension naturally does it as it cycles. It is a function of suspension geometry.

Do both the front and rear axles move the same amount at the same time?

They can, but it is unlikely.

 

[gt1guy]

The suspension naturally does it as it cycles. It is a function of suspension geometry.

They can, but it is unlikely.

If it does naturally, it's not much lateral movement at all. Ride height is generally pretty much as far out as each axle ends reach. Bump and droop both will follow an arc that has the axle ends move inboard. Even roll steer doesn't have the axle ends moving out more than they sit at ride height.

How do you plan to change the geometry to get the lateral movement up to inches?

If the lateral movement comes from geometry alone, it will absolutely be there in roll during a corner wouldn't it?

Or am I still missing what you want to do?

 

[Treefrog]

If it does naturally, it's not much lateral movement at all. Ride height is generally pretty much as far out as each axle ends reach. Bump and droop both will follow an arc that has the axle ends move inboard. Even roll steer doesn't have the axle ends moving out more than they sit at ride height.

If the travel is symmetric side to side, the axle ends do not move in or out.

In roll it moves inboard from the chassis perspective, but not the road's perspective. The road sees the body move side to side.

How do you plan to change the geometry to get the lateral movement up to inches?

A normal dual triangulated 4 link already does. It does seem that the closer the axle end of the uppers are, the move movement is has. A generic 4 link example:

If the lateral movement comes from geometry alone, it will absolutely be there in roll during a corner wouldn't it?

Yes. Corners, camber, a donut loving passenger.

Or am I still missing what you want to do?

I am not trying to do anything. I am just wondering about the trade off between shifting the axles back under the CG in roll vs minimizing lateral axle movement.

 

[Weasel]

What I am thinking is that you have two vehicles, A and B. A does not have any side to side movement of the center of the axles. Whereas for B, the center of the axles moves 3 inches from the chassis centerline towards the compressed side. Both have the same chassis and axle CGs. Vehicle A will roll over with less sideways Gs (from the axles' reference frame) than vehicle B.

got it yes, agree

Squatting, lifting, and torque roll are the result of drive forces not acceleration. Acceleration usually produces the highest force. There is force in the form of drag (air and driving surface) during steady state. Steady state uphill driving is another force example.

but those are not resolved via Anti's as in anti squat values? You could have a torque acting on the vehicle but without "A" there is no F=MA. Steady state uphill is "A" = 0.

 

[Treefrog]

but those are not resolved via Anti's as in anti squat values? You could have a torque acting on the vehicle but without "A" there is no F=MA. Steady state uphill is "A" = 0.

I am not entirely sure I understand what you are trying to say. Antis are just a guestimate of how the suspension reacts to drivetrain loading.

You are missing an important part of that equation. Sum(F) = MA. You do not have to be accelerating to have forces applied.

A quick example is a pen. You push down to click it, and hold it down. Your finger is putting force on the spring through the button. Your hand holding the pen is reacting to the force that is being transmitted through the spring. The pen does not accelerate but it still deforms from its not pressed state.

 

[Treefrog]

So, I realized that I made a mistake in how I interpreting and presenting the axle shift. I made it seem as though I was showing represented the wheels on the ground. This was not the case. I was incorrectly showing the movement of the point where the axle shaft line crosses the center plane. I reran the analysis looking at the movement of the point halfway between where the wheels touch the ground. Two of the plots are the same as before.

The biggest observation I have is that parallel uppers have less contact shift.

First up, flex steer:

Axle center shift:

And the new plot, and shift of the ground contact points:

 

[Weasel]

Guess I was thinking there is a diff between static and dynamic forces.

 

[Treefrog]

Guess I was thinking there is a diff between static and dynamic forces.

There kind of is. Unless you are dragging your buddies broken rig, the difference in forces is easily an order of magnitude. Unless it is a soft suspension with lots of movement and high squat geometry, the static may be very hard to tell apart from at rest.

A quick example with some guestimates. I believe I remember a comment from Jeep regarding why the Hurricane concept had cylinder deactivation saying that they had found that a car needed 20 hp to drive down a highway. Taking this to mean a stock Jeep (4000lb) at 70 mph needs 20 hp. Doing some unit conversion and math, I ended up with the equivalent of .02 gs of acceleration. Take that with a grain of salt, force and mass are annoying with units.