2-Piece Split Brake Disc Forces

[CraneEng87]

Hello,

I am currently working on a project a little outside my comfort zone and would like some assistance in recognizing forces and stress in a particular area of a Brake Disc design.

The brake disc gets mounted to a drum which is already installed and the disc must be made into 2-pieces to allow for installation. I have attached a basic drawing of the disc mounted to the drum. I have already assessed the drum flange to disc bolted connection which consists of (16) 1"-8 Bolts and have verified that only 8 bolts are needed to resist the braking forces caused by the caliper (not shown) this also includes a 5 to 1 safety factor.

The disc splice is an over lap with one side tapped and the other countersunk. My question relates to this area. What kind of forces will I see at the splice? Assuming the caliper will only grab one half of the disc at a time and each half has (8) bolts to the flange that can resist the forces. I assume that since the brake force will be directly translated to the flange bolts that very little of this force will be transferred through the splice to the other bolts. If this is true then the bolts at the splice are only needed to keep the two halves of the disc together.

I'd appreciate any additional thoughts on this assessment.

 

[btrueblood]

The disc transfers the load from the caliper to the drum via shear stresses, and so the bolts at the splice will need to transmit some amount of those stresses as the brake passes the splice.

 

[Tmoose]

Is this for routine dynamic stopping, or just holding when stopped?

 

[tbuelna]

The disc lap joint fasteners should be designed for a shear fit at the body. This will be difficult using tapped holes.

The two-piece disc with lap joints can work if the disc halves are assembled and match machined at the friction and attachment faces, and then installed as a match set. A slight angle at the leading edge of the lap joint at the friction face would also help.

 

[desertfox]

Hi CraneEng87

I think I agree with you that the bolts holding the splice joint will not see any great forces, all or most of the braking force will be taken in shear by the larger 16 off 1" bolts.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[tbuelna]

Another thing to consider is the thermal growth difference between the rotor friction surface and where it is fastened to the hub. Sometimes this disc/hub attachment uses a floating arrangement of hat bushings to allow some radial expansion/contraction motion due to uneven heating between the hub and rotor friction surfaces.

 

[3DDave]

It's an indeterminant stress distribution problem. You need to make some assumptions to simplify it, like which bolts will take the load from the disk to the drum. Or that the full braking load will be applied to one end of one disk section and ignore the capacity of the other disk section to carry any load.

One thing to consider is the potential for an offset step between one segment and the next. If the pads are relatively rigid they won't accommodate much.

 

[CraneEng87]

Thanks for all of the responses, the holiday kept me busy so I'm just getting back to this.

I'll try to breakdown my responses by person. I have also attached a picture of the brake disc splice.

btrueblood:
I have assessed the flange bolts in shear and with frictional forces between the disc and flange. In shear alone it takes approx. 10 bolts to withstand the maximum braking force. When accounting for frictional forces between the disc and flange using bolt tension and a nominal friction coefficient of u = 0.33 I find that it would take 2.3 times the braking force to cause any slip. I will have to review the condition of when the brake disc passes or lands on the splice.

Tmoose:
I should have explained this in my original write up. This disc brake is for emergency dynamic stopping only and will be required to hold the load when stopped. It will not be used to routinely stop anything and in fact annual testing should be the worse duty cycle this brake ever sees. This brake is designed to stop the drum from turning as fast as possible and a full rotation of the drum is not expected once the brake is engaged. Ideally it should stop almost instantly.

tbuelna:
As of now the disc will be machined on both the mating surfaces and the braking surfaces. The disc halves will be assembled for final machining of the braking surfaces. I have a small 0.125" chamfer at the edge of each half to prevent a lip from forming (not shown in the attached picture). Referencing my answer to Tmoose the brake is not routinely stopping a dynamic load and we estimate that temperature increase will be around 0.69K per stop with full load.

3DDave:
Your approach matches what I recently did to assess the bolts at the splice in a worst case scenario. I assumed that the flange bolts were loose resulting in 0 frictional forces between the disc and flange. Since it requires about 10 flange bolts in shear to withstand the braking force without this frictional force, additional force will need to pass though the splice. I assumed that the caliper grabbing one half of the disc at a time would transfer maximum load to the 8 bolts holding that half to the flange and the remaining force would have to pass through the splice in shear to reach the other disc half flange bolts.

I plan to use the approach I explained to 3DDave to assess this connection unless someone else has another idea. I will look into the condition of when the caliper grabs or passes the splice. I believe that a grad at the splice will not increase any stress at this connection, it will in fact increase the frictional forces between the disc halves at the splice. I will have to think about what happens when the caliper passes the splice.

Again, I appreciate the responses.

 

[Tmoose]

I think if the caliper happened to grab the rotor just after the loose splice, the reaction would be more like peeling off one segment, pivoting around the further bolt.

Machining an interlocking feature in the lap joint would provide some redundancy there.
Kind of like this, but without the wedge.

https://www.fhwa.dot.gov/publications/research/infrastructure/structures/04098/images/fig106.jpg

The short, flat head (?) screws aren't capable of much clamping, so may be susceptible to loosening.

 

[CraneEng87]

I like the idea of a mechanical interlock to reduce the shear force on the screws. The screws are being sized off of these calcs so they are not definite at this point, however due to tapped thickness of one disc half I have selected 1/2"-13 screws to start. The disc thickness is 30mm so at the splice the tapped side is only 15mm and since I need the bolt sunk a little at each end I have about 13mm of thickness for thread engagement.

The clamping force I'm using is based off of the "recommended seating torque" provided by the manufacturer which is 850 IN-LB and the following formula.

T/K x D = P
T = Torque (850 IN-LB)
K = Torque Coefficient (0.20 Unitless)
D = Bolt Nominal Diameter (0.50 IN)
P = Tension (LB)

It comes out to 8500 LBF per screw. The screws are F835 Alloy Steel Flat Countersunk Head Cap Screws. According to F835-04 it states the allowable tensile load to be 20,600 LB which is based off the maximum tensile stress of 145,000 psi and a stress area of 0.1419 in^2.

 

[BUGGAR]

Can you get cyclical forces such that you should consider fatigue?

 

[CraneEng87]

Buggar:

This is an emergency brake only and will be tested at no load once a year, so I do not believe cyclical forces are a concern. The brake is designed for full load and will only be used if a catastrophic failure occurs to the system or if a power failure occurs during operation.