Rocker Arm Design

[rjwoods77]

Hey All,

I just got all of my suspension modeled and now I am doing the front and rear shock actuation system. I dont really have any constraints except for the shocks, and a roughly narrowed down locations for rocker pivots and pullrod location. I spent alot fo time last night shooting at a running target and I am wondering if there is a better method(or some method for that matter) to go about doing this. Any general rule of thumbs or do I keep shooting until I get something close?

[rjwoods77]

Hey All,

I just got all of my suspension modeled and now I am doing the front and rear shock actuation system. I dont really have any constraints except for the shocks, and a roughly narrowed down locations for rocker pivots and pullrod location. I spent alot fo time last night shooting at a running target and I am wondering if there is a better method(or some method for that matter) to go about doing this. Any general rule of thumbs or do I keep shooting until I get something close?

[Denny Trimble]

http://fsae.com/eve/forums/a/tpc/f/125607348/m/7241...10087611#11610087611 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/72410967611/r/11610087611#11610087611)

Pay special attention to the "monthly macro" link in my 2nd or 3rd post.

[kwancho]

I'm using Mitchells, and it's REALLY easy to use. I redesigned our front bellcranks on Saturday, and through trial and error (a lot less than i expected), I got it right back to the MR I wanted with very little change over the suspension travel.

[Mark Bacchetti]

Yeah, rocker design is a bitch.

I have developed some spreadsheets from The Shock Absorber Handbook by John Dixon to get our IR in the ballpark. The book has a fairly theoretical approach on motion ratio calculations. It's not simple but it is the only resource I have found on the subject. Even with the spreadsheet, we still spend a lot of time on iterations (mainly due to chassis packaging issues).

-Mark
Cal Poly Pomona

[Schulberg J]

I like to model the suspension in 3d in Solidworks. I insert a push rod and constrain one end of the push rod to the lower ball joint and the upper end of the push rod to the area of where it will attach to the rocker on the chassis side. I leave the length of the push rod unconstrained. I then cycle the suspension through its travel and look at the change in length of the push rod. This gives me all the information I need to design the rocker to get the MR I desire.

[Greg H]

I used a method similar to Jason. I modeled the uprights, frame, a-arms, shocks, rockers, and pullrods in Solidworks. You can then easily move the upright and watch the measurement of shock travel change. You can easily modify the arm lengths and the angle between them in Solidworks. This reduced time to about 10-20 minutes for front and rear and went from using 3 or 4 different programs to just one.

[Denny Trimble]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Schulberg J:
I like to model the suspension in 3d in Solidworks. I insert a push rod and constrain one end of the push rod to the lower ball joint and the upper end of the push rod to the area of where it will attach to the rocker on the chassis side. I leave the length of the push rod unconstrained. I then cycle the suspension through its travel and look at the change in length of the push rod. This gives me all the information I need to design the rocker to get the MR I desire. </div></BLOCKQUOTE>

So, your rocker arm must have infinitely long lengths from pivot to pushrod and shock attachment points... or are you ignoring the nonlinear (rising/falling rate) effects of a rotating rocker arm?

[Schulberg J]

Denny,

Yes I am completely ignoring the nonlinear effects of the rocker during this step of design. It is also only a tool for initial sizing of the rocker. The way I look at the change in motion ratio through out travel, and consequently what is also the next step in rocker design is as follows.

I establish a referance plane in the plane of the rocker rotation offset the proper distance from the chassis so as the push/pull rod and rocker are located the right distance from the chassis.

I draw a rocker in this plane. The rocker is 2 lines intersecting each other and I fix the intersection point. I constrain the angle between the lines, their length, and attach the push/pull rod to one end and another line representing the damper to the other end. Next I constrain the push/pull rod length and the end of the "damper" not attached to the rocker to where it will be attached on the chassis.

I can then cycle the suspension and look at the change in length of the "damper" line and thus shock travel. This method takes into account nonlinear response due to the rocker and is very easy to make design adjustments to ie; changing rocker geometry, damper attachment point, actuation rod length, rocker mounting location etc. I can have Solidworks open in one window and Excel in another and quite quickly pump out a curve showing MR through out travel.

The reason I advocated the first step, which ignores the nonlinear effects, is because without any trial and error it gives you the rocker ratio you need to get your desired MR at ride height. Maintaining or deviating from this MR is just a matter changing the length of your rocker arms while maintaining this initial rocker ratio (assuming you want to maintian your initial MR at ride height, which I believe to be a good assumption).

[Denny Trimble]

Sounds like you've got it figured out. Have you seen the excel/solidworks macro to automatically step the input dimension and record the output dimensions? It's in the link I posted above.

I've modified it to suit our needs, but it sure beats manual mode.

[Omer]

hey
do you use the models in solid as only lines or as a complete bodies and moving them on the Assembly ?
i'm unaware of an option of moving just lines pls enlight me.

I wrote a Matlab program to give me the IR.

first i input all the dimensions (rocker dimensions,angle of rotation in dependence to a-arms size...)

and then i have a loop raising the upright to bump and lowering to rebound.

my goal is first to find the amount of rotation of the rocker,and from there find the amount of change on the bumper.
matlab is solving 3 equations,intersection of 2 spheres and a plane,there are 2 solutions and the program knows to choose the right one.

[Schulberg J]

Denny,
Have not tried it.

Omer,
I am using just lines, no solid bodies. I am moving stuff around with in a sketch. The suspension is in one sketch, the rocker and shock in another.

[Omer]

thanks Jason

i'll give it a go and compare it with my program.

[Omer]

hi Jason
I tried it out but didn't really succeded.
on what plane do you sketch the push rod?
how do you make sure your push rod connection to the lower a-arms is fixed?

[Mike Cook]

We spent like two hard months trying to get mechanical simulation to work in I-deas. It never worked. In about a week we built matlab programs to crunch all the numbers for us instead. It doesn't like as fancy but it sure does work. Ideas is the biggest piece of shit program I've used (however, it's only program I've used...)

[Schulberg J]

Omer,
Idealy you want the pushrod in its own 3d sketch so it is not overly constrained. You conecct one end to the rocker, and fix the rocker pivot point, and connect the other end to the lower ball joint (assuming you are doing pushrod). A simple constrain connection works. Hope that helps.

[Dan J]

I have been using Lotus Shark software for the suspension design on our 2006 car.

Having selected a rocker pivot location and a damper to body location a rough set of first points for the rockers for my desired rocker leverages and push(front)/pull(rear) rods was entered to establish the approximate damper stroke.

Now having a good guesstimate for the damper length at static ride height the iterative process for detail design of the points started, after 30 mins of iteration the change in motion ratio for both the front and rear suspension is &lt;1% over the whole suspension travel, the change being rising rate in bump and falling rate in droop.

I would thoroughly recommendthe software to any other team, http://www.lesoft.co.uk
for more info.

Incidentally, what sort of rocker leverages are people running? There are dire consequences in high lever rates if manufacture tolerance is high.

Our car is 1:1.2 Front and 1:1.6 rear.