Composite Sway-Bar Design and Manufacturing

[6543w2qddsf]

I am in the process of designing a composite sway bar for automotive applications, the objective is to reduce the weight of the part.

I have been told to look into two manufacturing techniques:
- Filament winding
- Pulltrusion

From what i undestand (i have little experience with composites) they differ in that Filament Winding needs a core for every part manufactured, wile with Pulltrusion you can use a single mold to produce an X number of parts.

Assuming the above is correct the bar manufactured through Filament winding would be Hollow, while the pulltruded one could be solid. Is that corect?

Which manufacturing technique is cheaper? Faster?
Which one will provide the best material propreties?
Will the shape of the composite bar differe from the steel equivalent?

Thank you in advance.
Fegizii

 

[Joshmo]

In general, pultruding is the lower cost manufacturing method because of the high output but this method provides lower performence properties. Additionally, oreintation of fibers is generally limnited to the direction of the pultrsion process. If the design can work with axial stiffness then pultrusion may be a cost effective choice.

Regarding filamwnt winding, the properties can be signifigantly better with the proper fiber and resin choices but can be alot more expensive because of the lower output. Additionally, the fiber direction probelm is the opposite of pultrusion with the limation of fiber direction being more hoop wrapped than axial. Yes greater stiffness can be achievd by increasing the crossectional area of the part with a foam core but hand placement of foam core during the wrapping process adds signifigant cost.

Therefore understanding the loadcases of the sway bar would really need to be understood before choosing the manufacturing method

 

[MechanicalChef]

When I hear sway bar, I imagine something similar to this:

or

The first image couldn't be done with pultrusion. Pultrusions are by a large majority straight. Curved pultrusion does exist, but it cannot perform anything that complex. (as far as I am aware, given the current state of the technology.) To wind something like that would require a 6-axis winder, and high-end winding software. So I'd stay away from that.

The second image would be a bit more reasonable. The ends would still need to be mostly metallic if you're trying to adapt this to an existing automotive application. (you might be able to RTM or hand layup those, but I digress) But you could create a straight center section with either pultrusion or filament winding.

Sway bars operate using torsion, so from a structural stand-point, filament winding would give you a stronger (read lighter) product.

This is because you would have the ability to perform helical winding, which is akin to the pattern you will see on a composite driveshaft. You can tailor the filaments to wrap at 45 degree angles around the mandrel. For purposes of simplicity, the mandrel could be a "fly-away" piece which stays embedded inside the wound filament shaft. The weight penalty would be minor.

You could also pultrude a straight section as well. Furthermore, pultrusions aren't limited to solid shapes. You can pultrude hollow tubes as well. The drawback to this has been mentioned as well - the fibers are aligned axially. Hence, they would be best suited to tensile loading, and wouldn't perform as well in torsion. Therefore, on an equal strength basis, it would weigh more than filament wound product.

For racing applications - I'd try and outsource wound shafts. For commercial sales, I'd 'consider' pultruding based on the production rate. You might not be able to get the desired torsional strength at the price of shedding vehicle weight.

I can't help you estimate capital equipment costs, unfortunately. I haven't been in the winding business for very long, and ours was over a million. Of course it's several orders of magnitude larger than what you'll need.

 

[Compositepro]

The only practical way I see to make such parts would be from braided tubes that are slipped over an inflatable mandrel which is then resin transfer molded. All the different shapes could be accomodated in a single molding. A braid can be expanded and contracted to make flanges and will go arround curves. It is also ideal for torsional load.
A&P Technology makes braids.

http://www.braider.com/index_hi.html

 

[tstanley]

Would it be possible for this application to use the braids over a foam rod type of mandrel? I've been looking for a source of foam rod for a similar application but have never found one.

Tom Stanley

 

[Compositepro]

A foam mandrel is theorectically possible but generally isn't practical when you consider the requirements it needs to have. It has to be sort and conformable to fit in the mold but then it needs to be non-porous, so it doesn't absorb resin, and not compress under molding pressure. That is why I suggested a bladder (which could be filled with foam).

 

[tstanley]

Thanks for your reply. I was afraid of that. What has been suggested to me is to use the bottles of expandable foam and use a long fairly flexible straw to fill the braided sleeve. What you say makes sense.

Tom Stanley

 

[Joshmo]

Perhaps use a closed cell syntactic foam in a seperate mold to form the core. Loctite makes a product called Synspand which is a synatactic film that expands 300 to 650% to fill a mold. Since its closed cell, there are no worries about resin saturating the core.

http://www.loctiteaero.com/Images/Datasheet_PDF/SynSpand_X9899CF.pdf

http://www.loctiteaero.com/Images/Datasheet_PDF/SynSpand_9899.pdf

The braid or dry fabric could then be wrapped around the core. Resin could be applied by a closed mold RTM process, bagged VARTM process, or wet lay-up...just an idea

 

[tstanley]

Thanks Joshmo. I wasn't aware of that product, so I gave you a star. That helps me in my application which is a similar problem to the sway bar.

I agree with Compositepro that braid is the way to go for the sway bar because it gets the fibers oriented best for torsion without the cost of the filament winding.

Tom Stanley

 

[InnovativeComposites]

Been there, done that.

A Sway bar works as a torsion spring. Fiber alignment must be in a 45° bias to provide function. Additionally, you can expect a modulus of around 15Msi with a uni-directional laminate from pre-preg materials. Filament wound UD fiber around 12-13Msi, Braid/RTM would be doing good at 9Msi, Pultrusion < 5Msi.

You are trying to replace steel so giving up quite a bit in material stiffness. You will need to get your "I & J" up to equal a smaller diameter steel bar. This can lead to size issues.

Built carbon bars for the factory Viper GTS cars several years ago. These where over 2" OD. Not an option on a street vehicle. Bar was manufactured by laying up uni-directional pre-preg carbon fiber over a bladder to the shape of the finished bar. This lay-up was mostly +/-45° through the center then some logitudinal fiber in the return ends. Once lay-up, placed in a female cavity mold, clamped in a heated press then the bladder was pressurized to consolidate the laminate against the mold walls. Worked great but was slow and expensive to build.

Hope this helps,

Steve

 

[racingcrazy]

I think you might find it quite hard to make a torsion bar with your required stiffness if the modulus of the materials involved is much higher than the metal that has been used.

Usually when I have had to design ARB's (american: swaybars), I have had to iterate with mmaterial, motion ratio's, wall thicknesses, diameters and lengths of the bar.

I try to achieve a desired stiffness at the wheel from the ARB, with the restriction that the wheel movement is finite. With this you play around with the dimensions and materials to make sure you have the desired stiffness without yielding the metal.

If you are looking for a straight swap, I think you may find it hard to get the ARB soft enough. i.e. generally a metal ARB will twist through a large angle with it's motion ratio to the wheel, and also the wheel generally tends to go through a large movement, even on open wheelers.