brakes revise [Archive] - FSAE.com Forums

Ben W

12-20-2011, 04:45 PM

I'll take a stab at this but just so you know, if you use proper grammar and punctuation, as well as better phrasing, it'll help others to understand and they will be more likely to answer your questions.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> pedal force it's obvious in rules the 2 KN force but i see it's too much high so built system on normal pedal force with pedal ratio 3 can be applied in all conditions and just used 2 KN to be sure the system will be safe and won't fail. </div></BLOCKQUOTE>

Not sure what you're saying / asking here. As I understand it, only the pedal needs to withstand 2kN, but you'd better make sure that the pedal box / lines / calipers / etc. don't break during the test. I decided that the required strength was too low for my comfort so I used a higher case load for wost case scenario and a more realistic one for fatigue and stiffness.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> 2-assumed deceleration i had will stop car from 100 Km/hr to zero i.e kinetic energy= 1/2*m*(100)^2 i think it's high too but took it for safety. K.E i took is high or normal? as usually won't stop from 100 km/hr suddenly in competition </div></BLOCKQUOTE>

You have discovered a rub that had bothered me as well... Let's re-phrase the problem in this manner: Assume the car loses no energy except from the braking system (no drag, energy into the tires, etc.) Make a series of acceleration runs with quick stops. There is a finite amount of energy that goes from the engine into the brakes and an average power. If you assume all of that energy is dissipated via convection you'll find that you'll need a very very high convection coefficient, high rotor area, or have your rotors run at very high temperatures.

Taking typical car values and about 90 square inches and 60 inches for the respective front and rear individual areas, the rotors would reach several thousand degrees F.

No the car doesn't go from 100 to 0 often, but the brakes are almost constantly being used. The 100 to 0 calculation is useful for evaluating how quickly your rotors come up to temperature. However, the energy from the engine running HAS to go somewhere, drag, tires, the air, the coolant system, and the rotors. Not all of it will go to the rotors but much of it will. Just how much depends on the track, among other things.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> 3-rotor we designed is from Cast iron and it's brittle material so factor safety need to be a bit high do you think 2 is enough? </div></BLOCKQUOTE>

Being brittle has more to do with what failure theory you use than the safety factor. I use a safety factor of one, but I also do the failure analysis for the absolute worst possible conditions for each rotor AND I designed the rotors to not fail after 120 hours with 50% confidence.

I was told we used a FoS of 3 before my time, but then again we've had one set of rotors fail because we analyzed the wrong material (and I am suspicious we neglected temperature effects).

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> 4-Temperature rise will affect on brake fluid and pad friction only or are there to take in consideration? </div></BLOCKQUOTE>

If something can have a temperature EVERY property it has is temperature dependent. Whether it is noticeable is another matter and is up to you as an engineer to decide.

I'd say from experience that the effect of temperature on the properties of brake fluid to not be noticeable from a performance standpoint, until it boils. Just keep your caliper temperatures below the fluid's boiling point.

Temperature DOES have an effect on friction between the brake pads and the rotors. Just how much depends on the compound, pressure, and sliding speed, among other things. Remember how we found the rotor temperatures after a series of stops? Rotor temperatures are more and more important as your pads are more and more temperature dependent. If you can keep your front and rear rotors the same temperatures, then you can pretty much ignore the temperature dependance of the pads, at least as far as front/rear balance is concerned.

-Ben

typeh

12-23-2011, 08:07 AM

sorry for language as i'm not a native English speaker.
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Not sure what you're saying / asking here. As I understand it, only the pedal needs to withstand 2kN, but you'd better make sure that the pedal box / lines / calipers / etc. don't break during the test. I decided that the required strength was too low for my comfort so I used a higher case load for wost case scenario and a more realistic one for fatigue and stiffness. </div></BLOCKQUOTE>
i meant to check safety for all components under 2000 N and don't build system on it .i mean not working on deceleration calculated from it. another point from your comment that you think it is low for comfort. i see it depends on driver and some teams use pedal ratio 1.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">You have discovered a rub that had bothered me as well... Let's re-phrase the problem in this manner: Assume the car loses no energy except from the braking system (no drag, energy into the tires, etc.) Make a series of acceleration runs with quick stops. There is a finite amount of energy that goes from the engine into the brakes and an average power. If you assume all of that energy is dissipated via convection you'll find that you'll need a very very high convection coefficient, high rotor area, or have your rotors run at very high temperatures. </div></BLOCKQUOTE>
i totally agree with you.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Being brittle has more to do with what failure theory you use than the safety factor. I use a safety factor of one, but I also do the failure analysis for the absolute worst possible conditions for each rotor AND I designed the rotors to not fail after 120 hours with 50% confidence. </div></BLOCKQUOTE>
my fear is to go in plastic zone so permanent deflection.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> If something can have a temperature EVERY property it has is temperature dependent. Whether it is noticeable is another matter and is up to you as an engineer to decide.

I'd say from experience that the effect of temperature on the properties of brake fluid to not be noticeable from a performance standpoint, until it boils. Just keep your caliper temperatures below the fluid's boiling point.

Temperature DOES have an effect on friction between the brake pads and the rotors. Just how much depends on the compound, pressure, and sliding speed, among other things. Remember how we found the rotor temperatures after a series of stops? Rotor temperatures are more and more important as your pads are more and more temperature dependent. If you can keep your front and rear rotors the same temperatures, then you can pretty much ignore the temperature dependance of the pads, at least as far as front/rear balance is concerned. </div></BLOCKQUOTE>
i meant boiling temperature. i appreciate your comment and opened my mind to new point (balance for heated rotors)

Thank you and i think this post will help others too.

AUM( Alexandria University Motorsports)
http://www.facebook.com/AUMotorsports

12-26-2011, 06:33 PM

typeh,

There are two main failure modes for brakes.
~~~o0o~~~

1. The brakes can break (ie. fall apart). The forces that do this come from two directions;

1.1. The driver pushing with 2+kN on the pedal.

1.2. The road pushing on the wheels. This is worst when the wheels have maximum vertical load and are on a high grip road surface, and can thus exert maximum torque on the disc, caliper, etc.

Because brakes are critical safety items, I would suggest you over-design for both these conditions. Bending is the preferable failure mechanism. "Breaking" is undesirable.
~~~o0o~~~

2. The brakes can overheat.

This is common in many racing series. You can tell when this is a problem because the brakes glow bright yellow at mid-corner.

Has anyone ever seen brakes glowing yellow hot in FSAE? Orange? Even dull red?

IMO overheating of brakes is not a big problem in FSAE. If it does become a problem you can use the same solution used on all other racecars;

Make a little scoop that catches the wind and throws it at the disc and caliper. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

Ben W

12-28-2011, 10:15 AM

I agree that overheating is not nearly as much of a problem in FSAE, but I've seen it happen (soft fade), and I am suspicious that hard fade has happened before on one of our cars.

If you have an old car to go off of, see what temperatures you are reaching and back out a convection coefficient. Otherwise, estimate one using heat transfer theory and estimate what will happen if you reduce the rotor surface area by X%. Keep your rotors happy and your fluid should stay happy as well. If you want to know fluid temperature stick a thermocouple somewhere on the a caliper, preferably away from a direct flow of air. It isn't exactly the temperature of the fluid, but it's close enough for an estimate.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> i meant to check safety for all components under 2000 N and don't build system on it .i mean not working on deceleration calculated from it. another point from your comment that you think it is low for comfort. i see it depends on driver and some teams use pedal ratio 1. </div></BLOCKQUOTE>

Pedal ratio of 1? I doubt it. That would require the master cylinders pointed at the driver's foot horizontally, or that the pedal would extend as far below the box as to the foot from the pivot point. Also, you need to know the mechanical advantage at the rotors AND the hydraulic advantage to know pedal effort, on top of pedal ratio and coefficient of friction.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> my fear is to go in plastic zone so permanent deflection. </div></BLOCKQUOTE>

If you design for service life, you won't see plastic deformation in the short term (if you did it right). If you don't want to go through the fatigue calculations, use a reasonable factor of safety.

-Ben

typeh

01-19-2012, 02:29 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
typeh,

There are two main failure modes for brakes.
~~~o0o~~~

1. The brakes can break (ie. fall apart). The forces that do this come from two directions;

1.1. The driver pushing with 2+kN on the pedal.

1.2. The road pushing on the wheels. This is worst when the wheels have maximum vertical load and are on a high grip road surface, and can thus exert maximum torque on the disc, caliper, etc.

Because brakes are critical safety items, I would suggest you over-design for both these conditions. Bending is the preferable failure mechanism. "Breaking" is undesirable.
~~~o0o~~~

2. The brakes can overheat.

This is common in many racing series. You can tell when this is a problem because the brakes glow bright yellow at mid-corner.

Has anyone ever seen brakes glowing yellow hot in FSAE? Orange? Even dull red?

IMO overheating of brakes is not a big problem in FSAE. If it does become a problem you can use the same solution used on all other racecars;

Make a little scoop that catches the wind and throws it at the disc and caliper. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z </div></BLOCKQUOTE>

Thanks Z
calculation done based on 2 KN and this is mechanism i can imagine and what i found in books i read about brakes that braking torque is highest load on rotor and must be designed on it.
i don't agree with 1.2 as loads fed into tires and from it to springs through push/pull rod and rotor almost only affected by torque on it

i totally agree with 2

Alexandria University Motorsports
FSG 2012

typeh

01-19-2012, 02:41 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Ben W:
I agree that overheating is not nearly as much of a problem in FSAE, but I've seen it happen (soft fade), and I am suspicious that hard fade has happened before on one of our cars.

If you have an old car to go off of, see what temperatures you are reaching and back out a convection coefficient. Otherwise, estimate one using heat transfer theory and estimate what will happen if you reduce the rotor surface area by X%. Keep your rotors happy and your fluid should stay happy as well. If you want to know fluid temperature stick a thermocouple somewhere on the a caliper, preferably away from a direct flow of air. It isn't exactly the temperature of the fluid, but it's close enough for an estimate.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> i meant to check safety for all components under 2000 N and don't build system on it .i mean not working on deceleration calculated from it. another point from your comment that you think it is low for comfort. i see it depends on driver and some teams use pedal ratio 1. </div></BLOCKQUOTE>

Pedal ratio of 1? I doubt it. That would require the master cylinders pointed at the driver's foot horizontally, or that the pedal would extend as far below the box as to the foot from the pivot point. Also, you need to know the mechanical advantage at the rotors AND the hydraulic advantage to know pedal effort, on top of pedal ratio and coefficient of friction.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> my fear is to go in plastic zone so permanent deflection. </div></BLOCKQUOTE>

If you design for service life, you won't see plastic deformation in the short term (if you did it right). If you don't want to go through the fatigue calculations, use a reasonable factor of safety.

-Ben </div></BLOCKQUOTE>
i found out the pedal ratio 1 here in forum during searching but i see it will be very hard on driver foot, hope someone from those teams explain their point of view here,i imagine they do it for the 2000 N in rules

for life what i studied in fatigue that part can fail on fatigue or static so normally check both and design part to withstand so if part will fail on static load it's totally independent on life or we consider load in brakes completely reversed load and design on fatigue only?

Alexandria University Motorsports
FSG 2012

Warpspeed

01-19-2012, 02:46 PM

I agree, red hot glowing discs and boiling brake fluid are not very likely in FASE, the kinetic energy and horsepower are just not there to do it.

But what is far more likely is pad fade caused by out gassing of fresh new pad material.

Bed your fresh new pads with stops of increasing severity, and really cook those pads. They will then be fine, until the bedded surface layer wears away with repeated wimpy braking.
Then when you do that really big stop, they could fade again, because the brakes have not been worked hard enough for quite a while.

Real race teams bed sets of brand new pads during testing, remove them, mark exactly where they came out from, put them back in the box, and keep them for the start of the big race.

Aggressive driving on brand new pads is just begging for brake fade.
Probably nothing at all wrong with the brake system design.
It's just something you need to know about, because it's a trap waiting for the unsuspecting.

JT A.

01-19-2012, 03:25 PM

We did some testing where we tried to make our rotors glow, just to see what it would take.

It ended up taking about 10 minutes of repeated 70mph to 0 braking tests, followed by a couple minutes of driving around at 20 mph using about 50% brakes and 50% throttle. Also, this was at night time. It was a very dull red that probably wouldn't have been visible in daylight.

TMichaels

01-19-2012, 04:16 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">This is common in many racing series. You can tell when this is a problem because the brakes glow bright yellow at mid-corner.

Has anyone ever seen brakes glowing yellow hot in FSAE? Orange? Even dull red? </div></BLOCKQUOTE>

Sure, FSE2010, night endurance, Uni Stuttgart. 94kW took their toll...:
http://media.formulastudent.de/FSG10/Hockenheim/20100807Saturday/2010080723-18-217288kroeger/1048870832_ZJUVs-X1.jpg

01-19-2012, 06:47 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
i don't agree with 1.2 as loads fed into tires and from it to springs through push/pull rod and rotor almost only affected by torque on it
</div></BLOCKQUOTE>
typeh,

The maximum rotor torque occurs when there is maximum vertical load on the wheel (friction - think about it). I would suggest calculating the brake's breaking load as;

Maximum-Torque = Wheel-Radius x ~5 x Total-Car-Weight!

Any less and I would not want to drive the car hard.
~~~~~o0o~~~~~

Tobias,

Does that car have brake scoops/ducts?

Again, the simple solution to brakes glowing red++, is to catch some of the wind and throw it at both sides of the disc and caliper. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

TMichaels

01-20-2012, 01:46 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Does that car have brake scoops/ducts?

Again, the simple solution to brakes glowing red++, is to catch some of the wind and throw it at both sides of the disc and caliper. </div></BLOCKQUOTE>

Yes, it has:
http://media.formulastudent.de/FSG10/Hockenheim/20100807Saturday/2010080715-51-012262almonat/1048880751_wxRoQ-XL.jpg

As far as I know the main problem was that they used their 2008 car as base design, including the brakes, but increased weight from 212kg to 278kg and of course reached higher speeds due to the increased amount of torque available.
This car was a beast with respect to torque. One of the most experienced drivers, Michael, seemed to have troubles controlling it even on the straights while he did demonstriation runs after the FSC2010 endurance. I have to say that I miss the unrestricted power of the FSE2010 cars quite a bit...

Ben W

01-20-2012, 09:09 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I agree, red hot glowing discs and boiling brake fluid are not very likely in FASE, the kinetic energy and horsepower are just not there to do it. </div></BLOCKQUOTE>

I'm not saying that it's not going to happen. If you don't have enough cooling to your brakes you WILL boil fluid. However, most teams usually have enough surface area on their rotors to do the trick. I remember Clemson saying that their brakes would soft fade during an endurance run in 2010. If you are worried about fade just here's a hint: we made our front rotors as large as possible in outer diameter (13" wheels) and made the OD to ID as wide as the pads with no through holes or slots, etc. We have yet to experience soft fade to my knowledge.

-Ben

01-20-2012, 11:44 PM

Tobias,

Wow, those are big scoops!

My main point is that it is very easy to get extra "-ve" horsepower out of your brakes - just add scoops. If only it was that easy to get more "+ve" power out of your engine!

An even better way to cure "brakes glowing red" is to add aero downforce. The faster you can go through the corners, the less you have to brake at the end of the straights.

Longitudinal G's = fun, but fuel expensive.
Lateral G's = better. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

typeh

01-22-2012, 03:10 AM

01-23-2012, 07:38 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">So 1.1 & 1.2 r same... </div></BLOCKQUOTE>
Typeh,

No, they are NOT the same. http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif
~~~o0o~~~

1.1 Driver pushes hard on the brake pedal -> The brake pedal bends or snaps, or the M/C pops out of its mounts, etc.

The rulemakers help you here by giving you a minimum pedal load of 2kN.
~~~o0o~~~

1.2 Road applies large torque to wheel during heavy braking -> The disc buckles, or shatters, or snaps off its mounts, or the caliper mount bends or breaks, etc.

Now, typeh, how big is this maximum expected wheel torque???????
Please give answer in kNm for a front wheel.

Z

(PS. There are other failure modes (eg. hydraulic), but let's get the easy stuff out of the way first.)

typeh

01-25-2012, 01:40 PM

i mixed coz i considered max. braking torque will be from max. pedal force so from my look it's one case but now i understand what u meant

front braking torque=279 N.m

01-25-2012, 08:18 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
front braking torque=279 N.m </div></BLOCKQUOTE>
Typeh,

If you design your discs/calipers/etc. for "ultimate failure load" of 279Nm, then they WILL fail.

That load corresponds to a horizontal braking force at one front wheelprint of about 120kg, which is easily exceeded.

Take front outer wheel vertical load during very hard braking and corner entry as, say, 200kg (because of load transfers). Now multiply by Cf ~1.5 (high grip tyre on high grip road), giving horizontal load of 300kg (could be Cf~2?). Finally, add a bump/pothole/kerb that increases vertical load, and thus also horizontal load, by, say, ... 3G? Could be more for a big pothole, but the organisers wouldn't do that, would they? Of course, you might be testing on that old parking lot out back...

Bottom line is that I wouldn't want the front brake structure failing at anything under wheel torque = 2kNm.

Or, to be on the safe side, 10x your 279Nm figure.

Would anyone who has done these calcs properly like to comment? My figures are plucked out of the air, and this is a fairly important safety issue.

Better yet, has anyone suffered catastrophic brake failure, and can they provide estimated failure loads?

Z

Buckingham

01-25-2012, 09:14 PM

Yes, I've seen brake rotor failure before in FSAE during normal testing, and it's a fatigue issue, not a single max load issue. (happened months after competition, guessing it has a few hundred miles on it).

Think about it... repeated thermal and mechanical stresses. It's not about if a rotor will fail, it's about when...

The crack also originated at a cross-drilled hole. Would I keep cross drilling, yes. It can be just as effective as a scoop when done properly, and weighs less.

Warpspeed

01-25-2012, 09:32 PM

Probably not a bad idea to adapt an existing commercial brake rotor off something or other, rather than trying to fabricate something totally from scratch.
After all, we don't attempt design our own calipers, wheels, or a whole host of other things.

Buckingham

01-25-2012, 10:35 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Probably not a bad idea to adapt an existing commercial brake rotor off something or other, rather than trying to fabricate something totally from scratch.
After all, we don't attempt design our own calipers, wheels, or a whole host of other things. </div></BLOCKQUOTE>

I disagree. Solid Disk Rotors are one of the cheapest and easiest parts on the car to design/make yourself.

can't say the same for calipers (still not hard to do, but harder to do better than premium off-the-shelf options).

Adambomb

01-25-2012, 11:06 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Buckingham:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Probably not a bad idea to adapt an existing commercial brake rotor off something or other, rather than trying to fabricate something totally from scratch.
After all, we don't attempt design our own calipers, wheels, or a whole host of other things. </div></BLOCKQUOTE>

I disagree. Solid Disk Rotors are one of the cheapest and easiest parts on the car to design/make yourself.

can't say the same for calipers (still not hard to do, but harder to do better than premium off-the-shelf options). </div></BLOCKQUOTE>

+1, rotors are a piece of cake if you don't overthink them too terribly (although this is FSAE...). On the other hand, I could spend hours massaging the rest of the parameters in the system to make it workable by adding the constraint of a rotor designed for another application that can't really be modified.

Warpspeed

01-25-2012, 11:28 PM

The reality is, that specialist manufacturers can mass produce a great many parts a lot cheaper, and almost certainly with much better quality control than you can do it yourself.

Engineering is just as much about practicality and cost saving, as it is about brilliance of design.

Nobody makes their own nuts and bolts, bearings, seals, rod ends, tires, or a great many other things.
It simply is not practical or cost effective to do so if "a bought one" is perfect for the job.

There are more than enough really challenging and vexing problems to tackle as it is.

typeh

01-28-2012, 04:49 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
front braking torque=279 N.m </div></BLOCKQUOTE>
Typeh,

If you design your discs/calipers/etc. for "ultimate failure load" of 279Nm, then they WILL fail.

That load corresponds to a horizontal braking force at one front wheelprint of about 120kg, which is easily exceeded.

Take front outer wheel vertical load during very hard braking and corner entry as, say, 200kg (because of load transfers). Now multiply by Cf ~1.5 (high grip tyre on high grip road), giving horizontal load of 300kg (could be Cf~2?). Finally, add a bump/pothole/kerb that increases vertical load, and thus also horizontal load, by, say, ... 3G? Could be more for a big pothole, but the organisers wouldn't do that, would they? Of course, you might be testing on that old parking lot out back...

Bottom line is that I wouldn't want the front brake structure failing at anything under wheel torque = 2kNm.

Or, to be on the safe side, 10x your 279Nm figure.

Would anyone who has done these calcs properly like to comment? My figures are plucked out of the air, and this is a fairly important safety issue.

Better yet, has anyone suffered catastrophic brake failure, and can they provide estimated failure loads?

Z </div></BLOCKQUOTE>

i didn't design accoording to this method my way start by assuming pedal force getting master cylinder pressure then force on peda from it get braking torque and decelration.
the way u r describing in brake handbook but i didn't feel it prefered this method.
i will try to compare results form both methods although seem great difference and will take in mind high factor of safety.
again u confirm to me heat load will be neglected compared to mechanical load http://fsae.com/groupee_common/emoticons/icon_smile.gif

thanks,

Alexandria University Motorsports
FSG 2012

typeh

01-28-2012, 04:54 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Buckingham:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Probably not a bad idea to adapt an existing commercial brake rotor off something or other, rather than trying to fabricate something totally from scratch.
After all, we don't attempt design our own calipers, wheels, or a whole host of other things. </div></BLOCKQUOTE>

I disagree. Solid Disk Rotors are one of the cheapest and easiest parts on the car to design/make yourself.

can't say the same for calipers (still not hard to do, but harder to do better than premium off-the-shelf options). </div></BLOCKQUOTE>

Accept with u and our team will try design our own calipers. FSAE is chance to learn so let's do it well and apply target from competition but still decisions need more wisdom and overall view.

Alexandria University Motorsports
FSG 2012

RollingCamel

01-28-2012, 11:33 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Buckingham:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Probably not a bad idea to adapt an existing commercial brake rotor off something or other, rather than trying to fabricate something totally from scratch.
After all, we don't attempt design our own calipers, wheels, or a whole host of other things. </div></BLOCKQUOTE>

I disagree. Solid Disk Rotors are one of the cheapest and easiest parts on the car to design/make yourself.

can't say the same for calipers (still not hard to do, but harder to do better than premium off-the-shelf options). </div></BLOCKQUOTE>

Accept with u and our team will try design our own calipers. FSAE is chance to learn so let's do it well and apply target from competition but still decisions need more wisdom and overall view.

Alexandria University Motorsports
FSG 2012 </div></BLOCKQUOTE>

Facepalm!

Really guys just focus on getting the car running well. Leave innovation for the coming years. At least if you bought most of the components and the car works then you will have a benchmark to equal or beat for the next car.

Design the calipers....good luck with that. Although it would be a good project for senior year teammates as team R&D, until proven to be working well.

Don't do the same mistake again...

TMichaels

01-29-2012, 03:53 AM

Just a comment from my side:
If you are designing your calipers on your own expect the scrutineers at FSG to put special emphasize on that and if they have any doubt in the presented solution, you will not pass.
I am not saying, that it is impossible, we had teams with self-designed calipers before, but they had a hard time in scrutineering and it became, more or less, the first part of the design report for them.

typeh

01-31-2012, 09:13 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TMichaels:
Just a comment from my side:
If you are designing your calipers on your own expect the scrutineers at FSG to put special emphasize on that and if they have any doubt in the presented solution, you will not pass.
I am not saying, that it is impossible, we had teams with self-designed calipers before, but they had a hard time in scrutineering and it became, more or less, the first part of the design report for them. </div></BLOCKQUOTE>

Thanks for appreciated advice http://fsae.com/groupee_common/emoticons/icon_smile.gif
maybe will be better to keep design on papers and discuss with judges about it and apply it next year.

AUMotorsports
FSG 2012

typeh

02-04-2012, 06:47 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
front braking torque=279 N.m </div></BLOCKQUOTE>
Typeh,

If you design your discs/calipers/etc. for "ultimate failure load" of 279Nm, then they WILL fail.

That load corresponds to a horizontal braking force at one front wheelprint of about 120kg, which is easily exceeded.

Take front outer wheel vertical load during very hard braking and corner entry as, say, 200kg (because of load transfers). Now multiply by Cf ~1.5 (high grip tyre on high grip road), giving horizontal load of 300kg (could be Cf~2?). Finally, add a bump/pothole/kerb that increases vertical load, and thus also horizontal load, by, say, ... 3G? Could be more for a big pothole, but the organisers wouldn't do that, would they? Of course, you might be testing on that old parking lot out back...

Bottom line is that I wouldn't want the front brake structure failing at anything under wheel torque = 2kNm.

Or, to be on the safe side, 10x your 279Nm figure.

Would anyone who has done these calcs properly like to comment? My figures are plucked out of the air, and this is a fairly important safety issue.

Better yet, has anyone suffered catastrophic brake failure, and can they provide estimated failure loads?

Z </div></BLOCKQUOTE>
According to method you state yes braking torque is near 2 KN.m .
vertical load on wheel due to max.braking + max.lateral acceleration + bump load.

I considered before there is no bump load ,also didn't take in consideration lateral load.

This torque is resulted from tire forces; so i calculate caliper pressure from it ,or just from pedal force? i think from pedal force because pressure generates due to force applied on pedal.

AUMotorsports
FSG 2012

flavorPacket

02-05-2012, 12:09 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TMichaels:
I am not saying, that it is impossible, we had teams with self-designed calipers before, but they had a hard time in scrutineering and it became, more or less, the first part of the design report for them. </div></BLOCKQUOTE>

If you follow a real design process and validate your parts, you should not have any issues. For example, Michigan has had self-designed and -produced calipers for more than 10 years and never had a problem with tech inspection at any competition.

On the other hand, frantically rebuilding a caliper in 10 minutes on the autox course at Silverstone because of a machining issue is not something you're likely to do if you buy a caliper off the shelf...

TMichaels

02-05-2012, 02:26 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">If you follow a real design process and validate your parts, you should not have any issues. For example, Michigan has had self-designed and -produced calipers for more than 10 years and never had a problem with tech inspection at any competition. </div></BLOCKQUOTE>

My comment was more directly related to their situation. Being a second year team with no remarkable experience and the obvious lack of resources, I would deem it to be more or less impossible and additionally unnecessary. Gains lay elsewhere in their situation.

flavorPacket

02-05-2012, 02:46 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TMichaels:
My comment was more directly related to their situation. Being a second year team with no remarkable experience and the obvious lack of resources, I would deem it to be more or less impossible and additionally unnecessary. Gains lay elsewhere in their situation. </div></BLOCKQUOTE>

Absolutely. This competition is about resource allocation and execution. Engineering is really the easy stuff!

PS I'm American and put my Walter Roehrl quote in German...funny that you put yours in English...

02-06-2012, 05:30 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
yes braking torque is near 2 KN.m .
vertical load on wheel due to max.braking + max.lateral acceleration + bump load.
...
This torque is resulted from tire forces; so i calculate caliper pressure from it ,or just from pedal force? i think from pedal force because pressure generates due to force applied on pedal.
</div></BLOCKQUOTE>
Typeh,

Yes, the caliper pressure is generated by the pedal force. But you also have to consider the various leverages involved (mechanical at pedal, and hydraulic between MC and calipers).

An important decision you have to make is this:
How much force must the driver apply to the pedal to lock-up the brakes?

If you use the "maximum pedal force" of 2kN (200kg), then the driver will need legs like an Olympic weightlifter. I suggest a brake lock-up force of about 0.5-1kN (50-100kg) for easier driving. (Anyone like to post their numbers?)

If lock-up occurs at 50kg pedal force, then from my previous calcs (on smooth track) this corresponds to front wheel torque of (roughly) ~300kgxGx0.25m = 0.75kNm. Now, if pedal force increases x4 to the maximum of 200kg, then potentially the wheel torque can also increase x4 to ~ 3kNm! It just depends how big that bump is...
~~~~~~o0o~~~~~~

I should also point out that if a structure fails after "a few hundred hours", which might be a few thousand cycles (Buckingham's example earlier), then it is not too far from failure by a single overload.

Putting it another way, "a few hundred hours" of moderate braking might be the same as ten hours of aggressive braking on a bumpy track. In this case the components should be "lifed" and exchanged after, say, 5 hours. But that's sounding like F1, not the "amateur weekend autocrosser". http://fsae.com/groupee_common/emoticons/icon_smile.gif

For brakes, I suggest a generous "safety factor".

Z

TMichaels

02-06-2012, 05:45 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Putting it another way, "a few hundred hours" of moderate braking might be the same as ten hours of aggressive braking on a bumpy track. In this case the components should be "lifed" and exchanged after, say, 5 hours. But that's sounding like F1, not the "amateur weekend autocrosser". </div></BLOCKQUOTE>

And since they intend to take part in FSG2012 and our head of dynamics really likes to put hard braking zones directly in bumps...

whiltebeitel

02-06-2012, 11:21 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
I suggest a brake lock-up force of about 0.5-1kN (50-100kg) for easier driving. (Anyone like to post their numbers?)

Z </div></BLOCKQUOTE>

We have used just under 0.6 kN pedal force for full braking with good driver feedback.

Der Krug

02-07-2012, 06:16 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Adambomb:

+1, rotors are a piece of cake if you don't overthink them too terribly (although this is FSAE...). On the other hand, I could spend hours massaging the rest of the parameters in the system to make it workable by adding the constraint of a rotor designed for another application that can't really be modified. </div></BLOCKQUOTE>
+2
We send out the DXF file for our rotors and get them cut out of 1020 for free.

Rotors are probably one of the easiest things to manufacture on an FSAE car if you have the right connections.

In my opinion the only reason for holes and slots are for slightly better rotor bite, slightly better cooling, and cosmetics.

In other words, rotors are not something to spend an excess amount of time on, compared to...everything else. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Warpspeed

02-07-2012, 07:36 PM

You won't find mild steel brake rotors on any production car, or any serious race car. Try ordering mild steel brake rotors from Girling or Brembo, they are just not made. Cast iron is a far superior material because it very rapidly work hardens, just like the bores of a very soft cast iron engine block become extremely hard.

Brake rotors are drilled or slotted, and sometimes the pads are slotted to release any trapped gas that may contribute to brake fade.
It helps in the same way that a tire tread helps grip in the wet.
http://image.hotrod.com/f/26723903/hrdp_1003_05+brake_pad_technology+.jpg

flavorPacket

02-07-2012, 10:18 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
sometimes the pads are slotted to release any trapped gas that may contribute to brake fade.
It helps in the same way that a tire tread helps grip in the wet. </div></BLOCKQUOTE>

Absolutely not. This was true when asbestos was used in friction materials. This is no longer the case, modern pads really only off-gas during the burnish and first fade.

Pads are slotted for many reasons including:

1) to change the natural frequency of the friction couple to eliminate/attenuate noise (squeal)

2) to reduce pad stiffness in order to improve the pressure distribution for better wear characteristics

Warpspeed

02-07-2012, 10:57 PM

Golly, I never knew race cars used drilled and slotted discs and pads to reduce brake squeal.

Usually the pad backing plate is made as stiff as practical to promote even all over pad pressure.
And slotting the pad material is going to have a negligible effect on total overall pad stiffness.

It is those ubiquitous thin metal stampings interposed between pad and piston that are used to damp resonances.

Even pad wear in very severe operation is usually achieved two ways, either by tapering the pads over their length, or piston stagger in multi piston calipers.

Cutting a giant slot through the middle is certainly not going to improve pad wear in any beneficial way, but it just might allow an escape path for trapped gas and debris.

flavorPacket

02-08-2012, 12:30 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
Golly, I never knew race cars used drilled and slotted discs and pads to reduce brake squeal.

Usually the pad backing plate is made as stiff as practical to promote even all over pad pressure.
And slotting the pad material is going to have a negligible effect on total overall pad stiffness.

It is those ubiquitous thin metal stampings interposed between pad and piston that are used to damp resonances.

Even pad wear in very severe operation is usually achieved two ways, either by tapering the pads over their length, or piston stagger in multi piston calipers.

Cutting a giant slot through the middle is certainly not going to improve pad wear in any beneficial way, but it just might allow an escape path for trapped gas and debris. </div></BLOCKQUOTE>

Hi Tony, it seems that you're explaining the racing version of brakes, where I'm trying to explain the OEM version of brakes. To me it's better to get students thinking about the entire system and all the dynamics involved. Focusing on the racing stuff only does students a disservice.

Please note I didn't say anything about rotors.

I work on brake NVH daily, I've got my data. The shims you refer to only work in the 5.5+kHz range, so often it's required to pursue structural changes to solve noise issues below this threshold. These often originate at the friction couple. Adding slots changes the couple which changes the frequency. Both CEA and brake ODS plots can show this.

If you feel otherwise, then we can just agree to disagree. It wouldn't be the first time people disagreed about brake engineering. But I would be remiss if I did not attempt to correct an inaccuracy.

I will give you one thing, pad slots can definitely be used to clear debris and let heat out. In racing, where people don't care about noise, it seems like a decent idea.

Warpspeed

02-08-2012, 12:52 AM

I certainly agree, braking requires a systems approach, there is a lot more to it than heaping together a bunch of individual parts.

How do you feel about mild steel disc rotors ?

The late Carol Smith had nothing good to say about mild steel rotors, and no OEM uses them at either end of the cost spectrum.

At the end of the day, an FSAE car is supposed to be designed as a purpose built race car, no judging points allotted for NVH.

Test Driver

02-08-2012, 09:52 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
You won't find mild steel brake rotors on any production car, or any serious race car. Try ordering mild steel brake rotors from Girling or Brembo, they are just not made. Cast iron is a far superior material because it very rapidly work hardens, just like the bores of a very soft cast iron engine block become extremely hard.

Brake rotors are drilled or slotted, and sometimes the pads are slotted to release any trapped gas that may contribute to brake fade.
It helps in the same way that a tire tread helps grip in the wet. </div></BLOCKQUOTE>

Tony, what application are the pads you pictured used on? I'm just curious.

flavorPacket

02-08-2012, 11:40 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
How do you feel about mild steel disc rotors ? </div></BLOCKQUOTE>

Steel will work, but there are plenty of good reasons why professionally designed rotors are not steel. Manufacturing, heat transfer, work hardening, the list goes on...

Warpspeed

02-08-2012, 02:45 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Test Driver:
Tony, what application are the pads you pictured used on? I'm just curious. </div></BLOCKQUOTE>
I have absolutely no idea.

If I wish to post a picture, to help illustrate an idea, I just go to "Google Image" type in some key words like "brake pad" or "three legged dog" and I get to choose something that seems most appropriate.
No way of knowing what that image actually is, or where it originally came from.

Der Krug

02-08-2012, 02:52 PM

How/why would you perform work hardening on brake rotors? AKA, plastic deformation..?

Warpspeed

02-08-2012, 03:01 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Der Krug:
How/why would you perform work hardening on brake rotors? AKA, plastic deformation..? </div></BLOCKQUOTE>
This happens anyway when you either start up a freshly built engine "and initially bed in the rings" or start using a fresh machined cast iron brake rotor.
The rubbing surface is placed under extreme shear stress, and that work hardens to a very hard smooth finish with superb wear characteristics.
Nothing to actually do. Just surface grind your cast iron discs true, fit them to the car, and the surface hardening process comes for free when you start using the brakes.
Mild steel does not do this.

Adambomb

02-08-2012, 04:17 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
The rubbing surface is placed under extreme shear stress, and that work hardens to a very hard smooth finish with superb wear characteristics.
...Mild steel does not do this. </div></BLOCKQUOTE>

And exactly how important is wear on our rotors? We have a car with something in the ballpark of 90 hours on it with no measurable rotor wear. I can also vouch that mild steel will bed in...after trading some parts around after a suspension failure it took some time to get the "new" side balanced to the "old" side. May not have as high of overall friction coefficient, but that can be accounted for by modifying your friction coefficient in your gain calcs.

I'm still not sold on cast iron in this application. When you can get laser cut steel for literally free (if you talk to the right people), it really doesn't make sense to spend the money on cast iron rotors, when the end result is you have to run maybe 50-100 psi higher line pressure and the system is still WELL within prescribed pressure limits.

So basically, given our limited resources the only real practical reason to run cast iron would be for some hand wavy appeasement of design judges. I can't justify it by engineering.

Warpspeed

02-08-2012, 04:43 PM

Fine, do it any way you want.

But just realize that over a hundred years of automotive engineering development have gone into steadily improving braking systems.
And if all the manufacturers are using a particular material or doing things in a particular way, then there may be some pretty compelling reasons why that is so.

Going directly against conventional engineering wisdom will usually lead to either fame or failure.

Adambomb

02-09-2012, 12:46 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
Fine, do it any way you want.

But just realize that over a hundred years of automotive engineering development have gone into steadily improving braking systems.
And if all the manufacturers are using a particular material or doing things in a particular way, then there may be some pretty compelling reasons why that is so.

Going directly against conventional engineering wisdom will usually lead to either fame or failure. </div></BLOCKQUOTE>

This isn't "going against conventional wisdom." This is "going against" common practice used on 3,000 lb passenger cars with unitized hubs and 2,500 lb race cars that spend a lot of time doing triple digit speeds. FSAE operates under a completely different set of constraints. Thermal constraints get superseded by torsional stiffness constraints, which later are superseded by cost and time constraints. It's about trading $500 in parts and even more labor for another set of tires and time to sufficiently destroy them. It's a different ballpark.

Warpspeed

02-09-2012, 12:54 AM

Deep in your heart, do you honestly believe that the braking system on an FSAE car is mechanically and thermodynamically totally different to any other braking application ?

TMichaels

02-09-2012, 02:00 AM

I am not a brake system guy, but I can confirm that my team has used laser cut brake discs for years with thicknesses which caused other teams to laugh at us and stating that we will never make it through endurance.
We proved them wrong every year including extra hard ABS testing and winning design.
And as far as I know we are still running the initial set of discs on our car from 2008 which has seen a lot of road under its tires.

Maybe other effects are more important in our application. Experience disproves theory in this case.

Zac

02-09-2012, 07:54 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Warpspeed:
Deep in your heart, do you honestly believe that the braking system on an FSAE car is mechanically and thermodynamically totally different to any other braking application ? </div></BLOCKQUOTE>

it is and it isn't. mechanically the system is the same, but the operating conditions and expected service life are completely different. Would a cast iron rotor be ideal? Probably. Is laser cutting brake rotors from steel good enough? Probably.

Mike Cook

02-09-2012, 07:55 AM

Flavor,

I would think that cast rotors are commonly used on larger race cars because they are very easy to cast with veins in them. This would be near impossible to manufacturer without casting. I'm certainly not a professional in the area of brakes, but reasons I can think of to go with one material over another are as follows:

1) Friction between pad and rotor
2) Strength at elevated temperature
3) Manufacturing
4) Cooling (convection and conduction)
5) Wear/endurance

I'm sure I missed some, but if you can comment on these points I would appreciate it. My gut feeling is:

1) Cast iron vs steel similar c.o.f.
2) Cast iron retains strength at elevated temperatures better than mild steel. However it is also much weaker to begin with. Even nodular steel I think is only about 35-40ksi at room temperature.
3) For FSAE our rotors are billet to begin with, and usually rough cut via waterjet. I don't see any reason to actually cast a brake rotor.
4) Not sure about cooling.
5) Due to the work hardening you mentioned of cast iron, it seems like they would last longer. Although I think this is not an issue on a fsae car at all. But it is a good point that I never knew. Thanks.

Anyways, I'm more curious than anything. For our team, we can buy mild steel plate in various thicknesses which make the manufacturing a lot easier than cast iron.

Also, does anyone have any good material properties for various materials at elevated temperatures?

Thanks,
Mike

Adambomb

02-09-2012, 02:12 PM

As a caveat, the main reason I'm even bringing this up is an attempt to "break the paradigm" so we don't have to spend as much time discussing this at design judging!

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> quote:
Originally posted by Warpspeed:
Deep in your heart, do you honestly believe that the braking system on an FSAE car is mechanically and thermodynamically totally different to any other braking application ?

it is and it isn't. mechanically the system is the same, but the operating conditions and expected service life are completely different. Would a cast iron rotor be ideal? Probably. Is laser cutting brake rotors from steel good enough? Probably. </div></BLOCKQUOTE>

Bingo. We don't go that fast, and our cars are very light. Also, I believe mild sheet steel rotors see successful use in motorcycles (not sure?), thinking I saw that somewhere. And again, thermal constraints allow for thin discs on a thickness magnitude comparable to sheet steel, which is cheap and easy to work with.

Our experience has shown that if you shoot for a target weight based on a given heat input and max temp, by the time you approach your target weight there is so little material left that you begin having stiffness issues in that it does not transfer braking torque through all the mounting points, resulting in very high contact stress at the floaters (necessitating much larger floaters and floater mounts, thereby offsetting any weight loss in the rotor). Therefore thermal constraints don't offer a real advantage one way or the other.

As for high temp strength, I don't have my source handy (had to do a bit of research to find that...), but as I recall the strength of mild steel at our target max temp was still comparable to that of cast iron. No real advantage here either.

As for c.o.f., I recall an SAE paper (again source not handy...) mentioning there was something on the order of at most a 10% difference, including surface effects. To me, that simply means you run a 10% higher pedal ratio! Lockup pressure goes from 600 psi to 660 psi (components all rated to 1500 psi...).

Bottom line is we've been running laser cut steel rotors since 2004, and have zero quantitative reason to change. Our 2010 rotors weighed something like 1.1 lb in the front, 0.7 lb in the rear. And again, even with a LOT of hours, we've never had to replace one. Regardless, we always get two sets made anyway, because this is the exact process we use:

1. Send .dxf of profile to Iowa laser
2. They provide material and cut them out of Grade 50 (similar to 1018) steel
3. Chamfer the speed holes and scuff the friction surface with some abrasives via a hand drill to aid in pad bedding, then clean (good job for a newbie)
4. File-fit floater mounts as necessary. (not necessary this year as the as-cut laser tolerances were within close to 0.001!)
5. Never think about them again! (except for doing temp measurements for iterative improvement...)

~~~~
Again, I bring this up in an effort to help break the paradigm that cast iron is ALWAYS better because "that's what real cars use" (ironically this seems to be one of a very small number of subjects where that statement gains any traction...). I know there are a good number of design judges that read this forum. In this application, given that all it takes to get in the top half is to just finish every event, and a very small number of teams have anywhere near enough testing on them to be really sorted out, there just aren't enough gains to be had to justify the cost and effort (Masturbation!!!).

I hope no other teams use this method! It would give us a nice competitive advantage. http://fsae.com/groupee_common/emoticons/icon_wink.gif I hope most teams continue go for the "ideal" solution of using cast iron for their rotors. And my money is that a good number of "top" teams will be risk averse enough to continue using cast iron (if they already are). There are probably a few forum lurkers on smaller teams that may heed this advice; more power to them. Although I intentionally left out my research...

Bottom line, as I've said repeatedly to our team, FSAE isn't about being the best so much as it is about just sucking less than the rest of the teams, which are ultimately staffed by overconfident amateurs. It's all about optimizing your compromises.

Kevin Hayward

02-09-2012, 05:33 PM

Mike,

You can get the cast iron in a billet form as well, which is generally the starting point for FSAE brake rotors.

It is generally easier to design steel brakes as the cast iron is horrible at dealing with tensile stresses. Looking at just the standard von mises stresses in your FEA can lead to cracks on the actual rotor.

The thermal and friction properties of the cast iron are significantly better than steel (of which I would recommend a 1030 or 1040). We used cast iron while I was at UWA and the guys have used steel at ECU. The guys have let me drive the ECU cars a couple of times and the braking system while very similar does not have as much initial bite and is maybe more temperature dependent.

That being said I would probably side with steel due to higher strength, and slightly easier manufacturing. Steel rotors are definitely good enough and can be made lighter than than cast iron.

I am surprised that the previous couple of posters are not suggesting running the rotors through a surface grinding process. Quite an easy manufacturing step and while it is time consuming can be run with little supervision. Making sure your discs start out flat improves your pad to disc contact.

Kev

Adambomb

02-09-2012, 10:10 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kevin Hayward:
I am surprised that the previous couple of posters are not suggesting running the rotors through a surface grinding process. Quite an easy manufacturing step and while it is time consuming can be run with little supervision. Making sure your discs start out flat improves your pad to disc contact.

Kev </div></BLOCKQUOTE>

Excellent point...we were looking into it this year, but couldn't quite get it hooked up.

02-10-2012, 04:06 AM

Typeh,

Well, you are certainly getting your brakes revised here. http://fsae.com/groupee_common/emoticons/icon_smile.gif

From the above it seems that "real racecars" only use cast iron brakes. Funny thing is that I remember back in 2005 much discussion about whether or not FSAE teams should start using carbon-carbon brakes. Because, after all, "that's what real racecars use"! http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

IMO, use steel.

Z

(PS. flavorPacket, what is your WR quote in English?)

TMichaels

02-10-2012, 04:16 AM

The quote basically says. "Good drivers have the fly residues on the side windows."

Sounds bumpy and is probably a bit too literally.

Mike Cook

02-10-2012, 06:49 AM

Thanks for the answer Kevin. Pretty much what I suspected. Let me give you an idea of our process for manufacturing rotors and this might shed some light on why I'd rather use mild steel.

1) Buy some nodular cast iron, I think it was from dura bar (http://www.dura-bar.com/downloads/upload/resourceGuide_06_appendices.pdf)

2) They only sell it in round stock, so you end up with a puck band saw on both sides and very oversized

3) Chuck it up in lathe face one side

4) Send to waterjet, rough cut

5) Chuck back up in lathe with crazy fixture to keep it from flexing, machine to .01" oversized

6) Send to surface grinder

7) Put on CNC and machine mounting surfaces.

The biggest pain in the ass is actually trying to machine it tough the oversized thickness on the lathe. You can't just chuck it up because it will not be stiff enough to maintain flatness. You need to have a backing plate with clamps, yada yada yada. It sucks. With mild steel, I can buy plates really close to our nominal thickness, waterjet it and grind it and post machine it fairly drama free. Sounds like a winner.

Mike

Kevin Hayward

02-10-2012, 09:36 AM

Mike,

As I posted I probably agree that steel is the way to go. Good enough performance, lower weight and manufacturing a bit easier. I think that the best advantage of steel is it is much more tolerant to design mistakes.

Both UWA and ECU have surface grinders in house. This means that you can be a bit lazier with the lathe and leave a bit of extra material if you are prepared to have the grinder run a little longer. You can also do some of the machining prior to making the individual discs and use the lathe to part them off. Depends largely on how complicated a pattern you use. With cast iron the discs tend to be simpler (and heavier) to avoid undesirable stresses.

One of the design goals of ECU has generally been to design parts that can be manually machined. This means that when a CNC becomes available (or more money arrives) the team can use it as a way to speed production rather than relying on the equipment. The mounting surfaces on the ECU rotors (steel & floating) weren't machined on a CNC.

If you desperately want to go cast iron and want to do it easy there is a US supplier selling ground cast iron pucks for about $95 a disc. They will machine to a pattern for a little more. It is a reasonable option to swap money for time. Another option is removing extra material from formula ford discs.

Overall what I would be looking for in an effective and easy to implement brake system for FSAE would be:

<UL TYPE=SQUARE>
<LI> Same master cylinder size for front and rear
<LI> Same calipers front and rear (with 2 on the rear) Plenty suitable options from the rear end of sports bikes.
<LI> If you go second hand on either the master cylinders of the calipers make sure you rebuild them using new seals.
<LI> Stiff and strong brake pedal
<LI> Same thickness rotor front and rear to allow for the same mounting system front and rear (generally floating discs)
<LI> Try to balance brakes on rotor diameter. This will probably lead to a thermal imbalance as well as a likely static offset on your balance bar. Neither of these are big issues. Temps tend to be pretty low and you should still be able to get it well within your normal balance bar adjustment range.
[/list]

A couple of things to watch out for:

- Bleeding the brakes first time will take a lot of time and fluid. Patience is required. Remember that air travels up, but may get stuck at connections. A bad bleed ruins even the best of designs.
- Get good at preparing brake hose ends. Buy extras for practice if you have to. It is surprisingly easy to make mistakes here.

One of the best parts to focus a bit of design and manufacturing effort is the balance bar. Any slop or compliance in the brake balance bar and its mounting will result in unpredictable brakes. Better to have less brake performance with predictable balance than the reverse. A lot of the off the shelf cheap components aren't that great. The expensive stuff is, however it is easy enough to replicate a good balance bar and make it in-house.

If you have extra money for the brake system I would put it into the master cylinders. The extra dollars usually buys you precision. On most nice masters the feed from the reservoirs is covered almost immediately upon the piston moving, resulting in braking pressure from very small movements. Brakes are meant to be modulated by force not by distance.

Another more controversial point is that there is nothing in the rules that says the pedals have to be adjustable. Ergonomics judges generally like to see adjustable pedals, but I have been involved (student and faculty advisor) with 4 design wins with no fore-aft adjustment in the pedals. I would take a solid mounted brake pedal with very little mounting compliance over adjustment almost every day. Especially with individually poured foam seats, which are only needed for autocross and endurance drivers.

Hope that helps the original poster, and given Z some more fuel.

Kev

flavorPacket

02-10-2012, 12:06 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
PS. flavorPacket, what is your WR quote in English? </div></BLOCKQUOTE>

Lots to discuss here, but I only have time today for the easy answer:

Good drivers have the bug guts on the side windows.

typeh

02-11-2012, 08:36 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
Typeh,

Well, you are certainly getting your brakes revised here. http://fsae.com/groupee_common/emoticons/icon_smile.gif

From the above it seems that "real racecars" only use cast iron brakes. Funny thing is that I remember back in 2005 much discussion about whether or not FSAE teams should start using carbon-carbon brakes. Because, after all, "that's what real racecars use"! http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

IMO, use steel.

Z

(PS. flavorPacket, what is your WR quote in English?) </div></BLOCKQUOTE>

I was really thinking about using steel instead of cast iron as it can withstand more, also i had constraints from caliper chosen (thickness and diameter for rotor)and only steel resulted in accepted results.

First i thought i'm mistaken as passenger car rotors are of cast iron http://fsae.com/groupee_common/emoticons/icon_smile.gif , but now i got answer before asking hahaha.

I would like to add another thing about cast iron. Cast iron absorbs heat and stores it so time will lead to increase in rotor temperature hence low coefficient of friction for pad and fade in brakes.

I would like to open discussion about aluminum rotors. As you mentioned FSAE has almost neglected thermal stresses. Also, aluminum has good thermal diffusivity.

---------------
AUMotorsports
FSG 2012
Design Team Head

Buckingham

02-11-2012, 11:06 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I would like to add another thing about cast iron. Cast iron absorbs heat and stores it so time will lead to increase in rotor temperature hence low coefficient of friction for pad and fade in brakes.
</div></BLOCKQUOTE>

How do you know that the material selection is a primary factor in determining mean operating temperature of the system? How do you know that this assumed increase in mean operating temperature will degrade pad performance?

Be very careful of "anecdotal engineering". Is your FSAE car a Tiger repellent because you don't see any Tigers near it?

typeh

02-11-2012, 12:40 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Buckingham:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I would like to add another thing about cast iron. Cast iron absorbs heat and stores it so time will lead to increase in rotor temperature hence low coefficient of friction for pad and fade in brakes.
</div></BLOCKQUOTE>

How do you know that the material selection is a primary factor in determining mean operating temperature of the system? How do you know that this assumed increase in mean operating temperature will degrade pad performance?

Be very careful of "anecdotal engineering". Is your FSAE car a Tiger repellent because you don't see any Tigers near it? </div></BLOCKQUOTE>

I read it before don't remember where but i tired it in simulation.
According to what i remember, made 2 identical rotors one of cast iron and other of aluminum weight was about 460 and 230 gram something like this. Resulted temperatures were 120 C for cast iron and 90 C for aluminum.
condition of simulation was to stop car weight 280 Kg from 100 km/hr to zero with a deceleration 1.6 g.

Adambomb

02-11-2012, 02:16 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
According to what i remember, made 2 identical rotors one of cast iron and other of aluminum weight was about 460 and 230 gram something like this. Resulted temperatures were 120 C for cast iron and 90 C for aluminum.
condition of simulation was to stop car weight 280 Kg from 100 km/hr to zero with a deceleration 1.6 g. </div></BLOCKQUOTE>

Do you only plan on doing one stop before the brakes cool again?

typeh

02-11-2012, 02:28 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Adambomb:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
According to what i remember, made 2 identical rotors one of cast iron and other of aluminum weight was about 460 and 230 gram something like this. Resulted temperatures were 120 C for cast iron and 90 C for aluminum.
condition of simulation was to stop car weight 280 Kg from 100 km/hr to zero with a deceleration 1.6 g. </div></BLOCKQUOTE>

Do you only plan on doing one stop before the brakes cool again? </div></BLOCKQUOTE>
of course not but i was explaining diffusivity and how simulation confirmed this idea.

Buckingham

02-12-2012, 12:05 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">According to what i remember, made 2 identical rotors one of cast iron and other of aluminum weight was about 460 and 230 gram something like this. Resulted temperatures were 120 C for cast iron and 90 C for aluminum.
condition of simulation was to stop car weight 280 Kg from 100 km/hr to zero with a deceleration 1.6 g. </div></BLOCKQUOTE>

In order to arrive at the absolute temperatures that you state (120C and 90C), how did you determine the amount of heat energy being transmitted away from the rotor to the air flowing over the rotor?

If you blow enough 30 C air over the rotor, the rotor will never heat up above 30 C, no matter what the material. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Zac

02-12-2012, 04:51 PM

02-12-2012, 06:24 PM

Here are some numbers I dug up in old notes;

Metal --------- Specific Heat (J/Kg.Deg) ------ Conductivity (W/m.Deg)
================================================== ====================
Aluminium ----- 900 (+/- 30) ------------------ 200 (+/- 50, depending on alloy)

Cast Iron ----- 470 --------------------------- 30

Pure Iron ----- 455 --------------------------- 80
Alloy Steel --- 460 --------------------------- 40
Stainless(304)- 460 --------------------------- 15

These are only approximate, and conductivity changes with temperature (most reduce, but stainless steel increases).
~~~~~o0o~~~~~

The "outlier" is aluminium. For a given mass and a given energy input (say KE of the car) it only increases by half the number of degrees of the iron alloys. It also conducts the heat away from the surface and into the bulk of the metal much more quickly, for lower surface temps.

BUT!, with lower surface temps the heat energy takes longer to leave the disc via radiation and convection (internal vents would help here). So after repeated stops the heat energy keeps building up in the disc. And then, as all small boys know, when the aluminium gets a little past warm, ....... it turns into jelly! http://fsae.com/groupee_common/emoticons/icon_frown.gif

And then there are issues of mu, pad compatibility, wear rate (do you hard anodize?), etc., etc....
~~~~~o0o~~~~~

Typeh,

So, as a beginning team, I really think you should look at the reasoning of Der Krug, Adambomb, Mike Cook, and Kevin Hayward on the previous pages. KISS!

"Optimilised" brakes will make diddly-squat difference to your dynamic points, compared with "adequate" steel brakes. However, the process of "optimilising" may well mean you end up with ZERO dynamic points ("Err, bummer, .... car not finished!" http://fsae.com/groupee_common/emoticons/icon_frown.gif).
~~~~~o0o~~~~~

On the other hand, if you are just after some "bling", then use nitrided stainless steel like many cafe-racer sports bikes. It is reasonably cheap and simple (similar to steel in above posts) but a lot SHINIER! http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

(Edit: Look here,Wilwood Rotors. (http://www.wilwood.com/Rotors/RotorList1.aspx)
No shortage of steel rotors here (mustn't be for "real" racecars??? http://fsae.com/groupee_common/emoticons/icon_smile.gif).
Stainless priced in proportion to bling (still quite cheap).
Read the "Note & Cautions" regarding aluminium rotors.
Geez, you can even get titanium (check the price!).
End Edit)

Warpspeed

02-12-2012, 07:01 PM

Shiny is definitely important.
Chrome plated brake rotors seem to be popular at the hotrod shows, so they must be good.

Jay Lawrence

02-12-2012, 07:58 PM

If you're serious about your bling, buy a sprotor for your rear end...

UoW tested aluminium rotors during their first year. They eventually yielded (dramatically) and are now on the wall of shame. Since then the team has typically used some version of steel similar to that used on modern bike rotors. I'm not sure why the standards for road cars are even being mentioned here, when in terms of the masses involved a bike is a closer comparison. My 2c

typeh

02-14-2012, 03:26 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
Here are some numbers I dug up in old notes;

Metal --------- Specific Heat (J/Kg.Deg) ------ Conductivity (W/m.Deg)
================================================== ====================
Aluminium ----- 900 (+/- 30) ------------------ 200 (+/- 50, depending on alloy)

Cast Iron ----- 470 --------------------------- 30

Pure Iron ----- 455 --------------------------- 80
Alloy Steel --- 460 --------------------------- 40
Stainless(304)- 460 --------------------------- 15

These are only approximate, and conductivity changes with temperature (most reduce, but stainless steel increases).
~~~~~o0o~~~~~

The "outlier" is aluminium. For a given mass and a given energy input (say KE of the car) it only increases by half the number of degrees of the iron alloys. It also conducts the heat away from the surface and into the bulk of the metal much more quickly, for lower surface temps.

BUT!, with lower surface temps the heat energy takes longer to leave the disc via radiation and convection (internal vents would help here). So after repeated stops the heat energy keeps building up in the disc. And then, as all small boys know, when the aluminium gets a little past warm, ....... it turns into jelly! http://fsae.com/groupee_common/emoticons/icon_frown.gif

And then there are issues of mu, pad compatibility, wear rate (do you hard anodize?), etc., etc....
~~~~~o0o~~~~~

Typeh,

So, as a beginning team, I really think you should look at the reasoning of Der Krug, Adambomb, Mike Cook, and Kevin Hayward on the previous pages. KISS!

"Optimilised" brakes will make diddly-squat difference to your dynamic points, compared with "adequate" steel brakes. However, the process of "optimilising" may well mean you end up with ZERO dynamic points ("Err, bummer, .... car not finished!" http://fsae.com/groupee_common/emoticons/icon_frown.gif).
~~~~~o0o~~~~~

On the other hand, if you are just after some "bling", then use nitrided stainless steel like many cafe-racer sports bikes. It is reasonably cheap and simple (similar to steel in above posts) but a lot SHINIER! http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

(Edit: Look here,Wilwood Rotors. (http://www.wilwood.com/Rotors/RotorList1.aspx)
No shortage of steel rotors here (mustn't be for "real" racecars??? http://fsae.com/groupee_common/emoticons/icon_smile.gif).
Stainless priced in proportion to bling (still quite cheap).
Read the "Note & Cautions" regarding aluminium rotors.
Geez, you can even get titanium (check the price!).
End Edit) </div></BLOCKQUOTE>
My problem with steel is weight :S
Due to economical reasons preferred last year caliper and it requires rotor minimum thickness 6.35 mm so steel will lead to huge weight.
I think we won't loss a lot of aluminum strength as we will work on low temperatures (Also you mentioned it previously) and if it occurs, high safety factor for maximum torque is 2.4 so i can loss strength and still in safe side.

I checked old posts about aluminum and saw few support it and some teams did well with it and intended to use steel instead http://fsae.com/groupee_common/emoticons/icon_frown.gif .

AUMotorsports
FSG 2012

Kevin Hayward

02-14-2012, 07:40 PM

Typeh,

There are pretty good reasons why Aluminium wasn't really discussed as a viable material for brake rotors and the main discussion was between iron and steel.

You have a situation where there are elevated temperatures, dimensional stability is important, and high cyclic loading. By using aluminium you may save 1-2kg maximum. This is for a safety critical part that will prevent you from running the comp if it fails.

My main questions would be:

- What is the target weight of your car, and is there any other place you could find 1-2kg?

- Are you sure you cant run a disc thinner than 6.5mm? my bet would be that a thinner disc may have problems once you wear a significant amount of pad material, but until then you would be okay. You don't tend to use pads quickly in FSAE so it should be okay.

Kev

typeh

02-16-2012, 03:33 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kevin Hayward:
Typeh,

There are pretty good reasons why Aluminium wasn't really discussed as a viable material for brake rotors and the main discussion was between iron and steel.

You have a situation where there are elevated temperatures, dimensional stability is important, and high cyclic loading. By using aluminium you may save 1-2kg maximum. This is for a safety critical part that will prevent you from running the comp if it fails.

My main questions would be:

- What is the target weight of your car, and is there any other place you could find 1-2kg?

- Are you sure you cant run a disc thinner than 6.5mm? my bet would be that a thinner disc may have problems once you wear a significant amount of pad material, but until then you would be okay. You don't tend to use pads quickly in FSAE so it should be okay.

Kev </div></BLOCKQUOTE>
We target 280 Kg with driver.
Minimum thickness for rotor 6.35 mm and maximum 9 mm. Another thing, pad is big :S .
weight save will be at least 3 kg

Adambomb

02-16-2012, 04:09 PM

+1 to Kevin, there are certainly real reasons why you don't see Al widely used in "real race cars" or even small race cars, motorcycles, or karts where you see steel rotors either. Here's a hint: Our team ran Al front rotors for one year and one year only (2003). Immediately after competition they were replaced with steel rotors and we've never looked back. A good thermal analysis will tell you why, looking at not only what number specific heat is, but how it translates into a part. Also note that Al doesn't just get soft at high temps like steel. Ever try to hot work Al? It usually just sort of "goes away" without warning. Sometimes with comical results, other times with tragic results.

Now, if your pad is so large, perhaps that's another parameter to consider changing? We lost a lot of rotor weight switching from Wilwood Dynalite calipers to PS-1s.

RollingCamel

02-19-2012, 01:48 PM

I've enjoyed reading this discussion. Typeh if you think that you main problem is weight think again. Safety, cost, reliability, machine-ability and ease of use comes way ahead in a 1st year team's priorities.

You need to understand what really the judges are looking for from a 1st year team. Being overweight is something expected to some degree and will decrease by the following years. Also, an incomplete or a broken car can be expected too.

Are you going to do a steel design too just in case plan A didn't work?

typeh

02-19-2012, 03:18 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RollingCamel:
I've enjoyed reading this discussion. Typeh if you think that you main problem is weight think again. Safety, cost, reliability, machine-ability and ease of use comes way ahead in a 1st year team's priorities.

You need to understand what really the judges are looking for from a 1st year team. Being overweight is something expected to some degree and will decrease by the following years. Also, an incomplete or a broken car can be expected too.

Are you going to do a steel design too just in case plan A didn't work? </div></BLOCKQUOTE>
I don't expect failure in rotor. i expect failure may occur in bobbins

RollingCamel

02-19-2012, 03:30 PM

typeh

02-21-2012, 02:35 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RollingCamel:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RollingCamel:
I've enjoyed reading this discussion. Typeh if you think that you main problem is weight think again. Safety, cost, reliability, machine-ability and ease of use comes way ahead in a 1st year team's priorities.

You need to understand what really the judges are looking for from a 1st year team. Being overweight is something expected to some degree and will decrease by the following years. Also, an incomplete or a broken car can be expected too.

Are you going to do a steel design too just in case plan A didn't work? </div></BLOCKQUOTE>
I don't expect failure in rotor. i expect failure may occur in bobbins </div></BLOCKQUOTE>

Ok...lets say what if it didn't perform? Plan for the worst and wish for the best.......something nearly every Egyptian team do the opposite. </div></BLOCKQUOTE>

Will increase thickness or change material according to faster solution.

Ben W

03-02-2012, 11:06 PM

I do have to take issue with this:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> The "outlier" is aluminium. For a given mass and a given energy input (say KE of the car) it only increases by half the number of degrees of the iron alloys. It also conducts the heat away from the surface and into the bulk of the metal much more quickly, for lower surface temps.

BUT!, with lower surface temps the heat energy takes longer to leave the disc via radiation and convection (internal vents would help here). </div></BLOCKQUOTE>

Let's take a look at the Biot Number, which describes whether you can use the lump capacitance method, i.e. whether the temperatures vary significantly within the material. People say that cast iron keeps the outer surface at a higher temperature. According to what source you use, a Biot Number less than 0.2 - 0.1 means that, at least for heat transfer calcs, you can assume that the temperature doesn't vary through the volume.

The Biot Number is defined as:

Bi = h*Lc/k

where,

Bi = Biot Number
h = convective transfer coefficient
k = thermal conductivity
Lc = characteristic length (typically, the volume divided by surface area).

I'll use our rotors as an example, which are cast iron (80-55-06 ductile cast iron).

Lc = 70cc / 581cm^2 = 0.121cm = 0.00121m
k = 24.2 W/m-k
h = 90 W/m^2-k (this is a highball estimate, I'll let you measure it / calculate it for yourself, but rest assured it's probably lower for most speeds)

So, Bi = 90 * 0.00121 / 24.2 = 0.0048 << 0.1

As far we care, these rotors are uniform in temperature across the thickness in the swept pad area. If you think our rotors are thin, double, no triple the thickness. You're still in the sub 0.1 region.

Some cast irons are better than most steels for brake rotors. Yes, they are easier to make vented rotors from, which is likely why most OEM rotors are cast iron, but it also can handle thermal stresses a lot better. Even though the temperature distribution is pretty close to even across the thickness of the rotor, it will vary quite a bit more radially.

03-03-2012, 04:44 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Ben W:
Let's take a look at the Biot Number...
...
the temperature distribution is pretty close to even across the thickness of the rotor... </div></BLOCKQUOTE>
Ben,

Nice to see a thoughtful post with some engineering calcs in it. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Now, I'm not really a brake person (I reckon thick wooden sandals and a hole in the floor should do), but...
I see a steady-state analysis of a fundamentally unsteady problem.

Any comments from the real brake gurus???

Z

Buckingham

03-03-2012, 10:12 AM

Ben:

I've measured Overall Heat Transfer Coefficient (commonly 'U') of between 400 and 1400 [W/m^2-k] in FSAE (calculated 3% uncertainty). This number is HIGHLY dependent on rotor size and geometry.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">
It also conducts the heat away from the surface and into the bulk of the metal much more quickly, for lower surface temps.

BUT!, with lower surface temps the heat energy takes longer to leave the disc via radiation and convection (internal vents would help here). So after repeated stops the heat energy keeps building up in the disc. And then, as all small boys know, when the aluminium gets a little past warm, ....... it turns into jelly! </div></BLOCKQUOTE>

Z: this is poor 'anecdotal engineering' on so many levels that I don't know where to begin... The system behaves much more steady state than you are giving it credit for. After a very short number of corners/laps the system reaches quasi-thermal equilibrium where the rotor temp vs time graph looks like a smoothed sawtooth (temperature rise at the braking zones followed by slower temperature decay in between). In FSAE i've seen the amplitude (max-min over the course of a lap) of this sawtooth be as little as (20-75 deg K) measuring rotor surface temp with non-contact IR sensor. The sawtooth amplitude is very much dependent on 'U'.

Once the system has reached quasi-equilibrium and we can basically say we have a mean operating temperature, then the rate of energy storage essentially becomes zero.

Energy Rate In = Energy Rate Stored + Energy Rate Out

as we reach quasi-equilibrium (after a few corners/laps):

Energy Rate In = Energy Rate Out = U*Area*(Trotor - Tambient)

and since U, Area, and Tambient are constants we can solve for Trotor (the mean operating temperature) for a given Energy Rate In (per lap).

Yes, when we change U we will change our mean operating temp, but the PRIMARY mode of heat transfer in this application is the forced convection of air flowing over the rotor. The rate of conduciton of heat into the rotor, through the hardware, into the hat, into the suspension, into the chassis and then into the air blowing over the chassis is negligible.

Ben W

03-03-2012, 10:52 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">
I've measured Overall Heat Transfer Coefficient (commonly 'U') of between 400 and 1400 [W/m^2-k] in FSAE (calculated 3% uncertainty). This number is HIGHLY dependent on rotor size and geometry. </div></BLOCKQUOTE>

We've measured convection coefficients 10 to 25 times smaller than what you're posted. If you do the calculations for a flat plate with forced turbulent convection, at 80 mph (which is the highest that an FSAE car is supposed to get at endurance) is around 90 W/m^2-k. Now assuming that a rotor is a flat plate might be inaccurate, but even when I modified our brake temperature simulation program to use the calculated coefficients (that varies with airspeed and temperature) the front and rear temps stayed really close (like within 20 degrees F) to the measured values.

http://i116.photobucket.com/albums/o26/go_crazy4_free/h_plot.jpg

Yes, it does depend on rotor geometry (the longer the "plate, or bigger the rotor, the lower the average convection coefficient.), but the flat plate assumption got us within an order of magnitude of our average coefficient and matched our recorded data.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> The system behaves much more steady state than you are giving it credit for. After a very short number of corners/laps the system reaches quasi-thermal equilibrium where the rotor temp vs time graph looks like a smoothed sawtooth (temperature rise at the braking zones followed by slower temperature decay in between). </div></BLOCKQUOTE>

I agree. The amplitude and mean depend on the convection coefficient while the thermal capacitance of the rotor only affects the amplitude (i.e. the lower the capacitance, the quicker the rotors reach quasi-steady state).

-Ben

Ben W

03-03-2012, 01:52 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Now, I'm not really a brake person (I reckon thick wooden sandals and a hole in the floor should do), but...
I see a steady-state analysis of a fundamentally unsteady problem </div></BLOCKQUOTE>

Even when the car is just starting (i.e. when the system has NOT reached steady state) the Biot number still can indicate whether we can assume uniform temperature throughout the thickness. The Biot number becomes irrelevant when we reach steady-state, it just describes how accurate it is to model a thermal mass as a single, uniform capacitor (and how much do capacitors matter in DC analysis? http://fsae.com/groupee_common/emoticons/icon_smile.gif) .

-Ben

03-04-2012, 06:29 AM

Ben, Buckingham,

All these computers, and still the "quasi-steady" analyses. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Here is the original quote I was responding to:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by typeh:
According to what i remember, made 2 identical rotors one of cast iron and other of aluminum weight was about 460 and 230 gram something like this.
Resulted temperatures were 120 C for cast iron and 90 C for aluminum.
condition of simulation was to stop car weight 280 Kg from 100 km/hr to zero with a deceleration 1.6 g. </div></BLOCKQUOTE>
That does NOT sound very steady-state to me!

I then considered what would happen if this was repeated a few times, up until Typeh's lightweight aluminium rotor turned into jelly, most likely well before quasi-steadiness.
~~~o0o~~~

Anyway, I'd be interested to hear if anyone has done a detailed unsteady simulation (or real test?) of the above situation. That is, comparing cast iron and aluminium rotors, and giving the varying temperature profiles through the rotors.

To repeat, rotors start cold (ambient T), get short but intense frictional heating of surface, with simultaneous conduction, convection, and radiation, and then whole process repeated after short car acceleration period back to 100kph.
~~~o0o~~~

Also, in above posts I don't see much mention of radiated heat flows. Obviously, the hotter the rotor (black, red, yellow...), and for same wind speed, the greater the ratio of Q.radiation/Q.convection.

Can anyone give some rough numbers for the Q.rad?

Z

Ben W

03-04-2012, 06:15 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Anyway, I'd be interested to hear if anyone has done a detailed unsteady simulation (or real test?) of the above situation. That is, comparing cast iron and aluminium rotors, and giving the varying temperature profiles through the rotors.

To repeat, rotors start cold (ambient T), get short but intense frictional heating of surface, with simultaneous conduction, convection, and radiation, and then whole process repeated after short car acceleration period back to 100kph. </div></BLOCKQUOTE>

I have done this analysis with a material that is supposed to have a high temperature difference from the outer surface to the core (cast iron) and the temperature from the surface to the core varies less than 10 degrees F. Aluminum will, of course, have less of a temperature gradient, but both are negligible as far as transient heat transfer analysis is concerned. I never bothered going through the same test with aluminum because the cast iron alloy we chose had the highest resistance to thermal stresses of any machinable material in our library with an acceptable strength. Our rotors see over well over 30ksi and fatigue is a huge concern for us. Aluminum's strength is quite dependent on temperature AND it's not known for having very good fatigue properties. Add to that I couldn't find a good S-N curve for the "high-temp" aluminum alloys we'd even consider. The cast iron I chose had a good empirical correlation between stress, fatigue, and temperature.

Remember, our rotors are NOT load cycled every time we hit the brakes, they are cycled every rotation under braking. They are also cycled after the first heavy braking zone (first of the day), after each braking zone, and when the car cools down (i.e. driver change). So let me ask you a question, do you think a material not known for good "at-temp" strength or good fatigue life, should be trusted for this? Even if we tried to make it work, I'd wager that 80-55-06 cast iron has a better specific strength than aluminum at ~600 degrees. If you try to give your Al rotors a fighting chance and lower the operating temperature for your aluminum rotors, you have two options: increase your convection coefficient (hard to do) or increase your surface area. But if you increase your surface area, you increase weight, and you're right back to square 1.

I will repeat what I've heard many others say: "serious racers don't have aluminum rotors." That's not quite true, for instance mini-Baja almost never uses their brakes, as well as dirt sprint cars and other dirt oval racers, but any vehicle that has to endure repeated, heavy stops should use cast iron or the like.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Also, in above posts I don't see much mention of radiated heat flows. Obviously, the hotter the rotor (black, red, yellow...), and for same wind speed, the greater the ratio of Q.radiation/Q.convection.

Can anyone give some rough numbers for the Q.rad? </div></BLOCKQUOTE>

Let's assume that our rotors are away from our wheels or anything else that is shiny. In other words, that none of the radiation will be reflected back to the rotors (to calculate radiation you have to count how many ways the body can "see" itself.) and is surrounded by a black-body. Let us also assume a fairly high emissivity of 0.9.

Radiative power is given as:

P = e * SB * A * (T_o^4 - T_inf^4)

where:
P = heat transfer rate
e = emissivity
SB = Stefan-Boltsman constant
T_o = surface temperature
T_inf = background temperature
A = rotor surface area

We'll say our rotors are getting very hot, not glowing, which I've never personally seen in FSAE (not saying it's not happened, just that it's rare), but still quite hot. [ All math is in metric]

so P = 0.9 * 5.67*10^-8 * A * (375^4 - 22^4)
[P_radiation/A] = 1.01 kW / m^2

Convection: P = h * A * (T_o - T_inf)
P = ~40 * (375 - 22) * A
[P_convection/A] = 14.1 kW / m^2

Even with poor convection and good radiation, radiation only accounts for around 10% of the heat transferred to the surroundings and can be included in the convection coefficient.

Conduction is quite small (by design) but that can be rolled into the heat transfer coefficient as well.

Gruntguru

03-04-2012, 07:08 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Ben W:

I have done this analysis with a material that is supposed to have a high temperature difference from the outer surface to the core (cast iron) and the temperature from the surface to the core varies less than 10 degrees F. Aluminum will, of course, have less of a temperature gradient, but both are negligible as far as transient heat transfer analysis is concerned. </div></BLOCKQUOTE>
There are numerous videos (eg http://www.youtube.com/watch?v=5_B-zFwE2r0 (http://www.youtube.com/watch?v=5_B-zFwE2r0)of)) of glowing rotors that demonstrate the highly transient nature od rotor surface temperature. It is clear to me that convective and especially radiative transfers are orders of magnitude higher during and shortly after the braking event. It is also clear that the sharply elevated temperature is confined to the surface of the disc (the glow diminishes too rapidly to be more than surface deep).

In summary, the temperature diffence from surface to core is much higher than the 10 deg shown by your simulation. Convective and radiative transfers are much higher during this period. The question that remains is "is conduction (to the core) significant enough during this period, to make higher conductivity materials eg Al significantly less effective at rejecting heat to the environment?"

03-05-2012, 04:18 AM

Ben,

Again, thanks for bringing some engineering to the forum.

A few points regarding your above post:
~~~o0o~~~

1. I note that it was Typeh (the thread initiator) who proposed using aluminium for his rotors. Most responses, including mine, advised against this for reasons similar to yours (ie. it turns into jelly when hot).
~~~o0o~~~

2. For radiative heat flow you have:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">We'll say our rotors are getting very hot, not glowing, which I've never personally seen in FSAE (not saying it's not happened, just that it's rare), but still quite hot. [ All math is in metric]

so P = 0.9 * 5.67*10^-8 * A * (375^4 - 22^4)
[P_radiation/A] = 1.01 kW / m^2 </div></BLOCKQUOTE>
This is the same equation I would use.....
except.....
I would have the temperatures in Kelvin. So, say 650K rotor, and 300K ambient (rather than 375 and 22).

This gives a figure of about
P.rad/A = 8.7 kW/m^2,
so much closer to your figure for convection (=~14 kW/m^2).

The hotter the rotor surface, the more significant is radiation (because of fourth power).
~~~o0o~~~

3. I still reckon that during and just after that first brake application from 100kph (ie. the original question) the temperature gradient between surface and core will be much higher than 10degF. Of course, this depends on lots of things, such as how long after, how thick is the disc, etc. If I get some time this week I will try a few calcs myself, in an attempt to answer the question at the end of Gruntguru's post....

Z

Ben W

03-05-2012, 11:35 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> This is the same equation I would use.....
except.....
I would have the temperatures in Kelvin. So, say 650K rotor, and 300K ambient (rather than 375 and 22).

This gives a figure of about
P.rad/A = 8.7 kW/m^2,
so much closer to your figure for convection (=~14 kW/m^2). </div></BLOCKQUOTE>

I can't believe I made that mistake http://fsae.com/groupee_common/emoticons/icon_redface.gif.

If you believe your rotors are getting hot enough for radiation to be important (I just recalculated ours and it's around 30% of the total heat rejected) you can linearize the radiation equation (where "h" is dependent on temperature), but it's probably just easier to do FEA at that point.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I still reckon that during and just after that first brake application from 100kph (ie. the original question) the temperature gradient between surface and core will be much higher than 10degF. Of course, this depends on lots of things, such as how long after, how thick is the disc, etc </div></BLOCKQUOTE>

Thickness of the disc does have an impact on the temperature gradient, but even when I put radiation into our ANSYS analysis we still didn't see much of a gradient.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> It is also clear that the sharply elevated temperature is confined to the surface of the disc (the glow diminishes too rapidly to be more than surface deep). </div></BLOCKQUOTE>

How quickly something cools down is not a direct indication of interior temperatures. Let's assume that you are correct, that the thermal energy pretty much stays at the surface. An otherwise identical rotor, except with a lower heat capacitance and higher conductivity will look the same. It will glow red hot and cool down very quickly. But it will have a hot core as compared to the other rotor.

You can't just look at the surface temperatures to determine the core temps. You can calculate the core temps if you know the thickness, thermal resistance, and heat capacity. Those same properties are used in computing the Biot number, which tells you which is of greater effect: the conductance inside the body, or at the surface.

Let's say the heat transfer coefficient is as high as 1400 W/m^2-k.

Bi = h*Lc/k
Bi = 1400*0.00121/24.2
Bi = 0.07

So for our fairly thin rotors, the temperature gradient is negligible. If you have rotors twice as thick or somewhat smaller in surface area (or both), then you can't really say that the gradient is negligible. However, I've never seen a heat transfer coefficient that high, even including radiation (which I stand corrected on, it is noticeable even for non-glowing rotors). I'd wager a guess that most rotors fall in the sub 0.1 Biot number region.

03-08-2012, 04:35 AM

I haven't finished the following yet (not enough room in the margins of the TV times http://fsae.com/groupee_common/emoticons/icon_smile.gif), but here is a start to some transient brake heating calcs.
(Note most calcs done by hand so lots of rounding.)
~~~o0o~~~

Assume a 280kg car stopping at 1.6G from 100kph (as per Typeh's original example). Also assume that between stops the car accelerates at 0.8G back up to 100kph, then repeats the cycle.

So maximum velocity, V = 28m/s, and
kinetic energy, KE = 1/2 x 280 x 28^2 = 110 kJ.

Deceleration rate, Ab = 1.6G = 16m/s^2
so braking time, Tb = 28/16 = 1.75s.

Acceleration time is simply double this ('cos half the Gs), so Ta = 3.5s.

Also total decelerating force, Ft (= m.A) = 280 x 16 = 4.48kN
but since aero and tyre drag will contribute to this, let's say that decelerating
force from brakes Fb = 4kN.

Therefore, power going into brakes initially is
Pb = Fb.V = -4000 x 28 = -112kW.

This power drops linearly with velocity to zero at 1.75s giving
energy absorbed/dissipated by brakes, Eb = 1/2 x 11200 x 1.75 = 98kJ,
but let's say Eb = 100kJ, (ie. a bit less than the KE, and rounder http://fsae.com/groupee_common/emoticons/icon_smile.gif).

~~~o0o~~~

Now let's assume the car has two sets of discs available, one of high conductivity aluminium, the other of very low conductivity stainless steel (or a special low-K cast iron). The aim is to compare the transient heat profiles due to the two different conductivities.

Each set consists of four discs, all equally loaded during braking. Specs are:

Disc OD = 160mm
Disc ID = 100mm
Disc Area (total 8x faces) = 0.1m^2

Property -------------------- (Units) --------- Aluminium ----- Stainless Steel
================================================== =====================
Density, Rho --------------- (kg/m^3) ------ 2700 ----------------- 7900
Specific Heat Capacity, C - (J/kg.K) ------- 900 ------------------- 460
Conductivity, K ----------- (W/m.K) ------- 200 ------------------- 15
Diffusivity, K/(C.Rho) --- (m^2/s) -------- 82e-6 ----------------- 4e-6
Thickness, Ld -------------- (m) ----------- 0.012 ---------------- 0.004
Mass (each disc), Md ------ (kg) ------------- 0.4 ------------------ 0.4

~~~o0o~~~

Some observations:

1. Worth repeating that these are just rough calcs, order of magnitude stuff, to get a feel for what is important.

2. The two sets of discs (Al and SS) are the same mass to try to keep the comparison to the K's. However, this makes the Al discs 3x thicker, which in practice means that they won't buckle so soon. Anyway.....

3. If all the braking energy is dumped into the discs (none into the pads), with no convection or radiation, and the energy is distributed evenly through the discs, then the increases in temperature are;

Aluminium delta T = Eb/(C x 4.Md) = 100,000/(900x1.6) = ~70 deg C (or K)
St.Steel delta T = = 100,000/(460x1.6) = ~140 deg C.

This is the upper limit to temp increase from each stop.

4. From above it is apparent that it will take a few more such stops before the disc core temps level off to anything approaching "quasi-steady-state". However, the SS discs may get up to temp faster due to their greater delta Ts ('cos lower heat capacity C), although this also depends on the relative rates of heat dissipation.....

5. An important number is the power input to the discs. This peaks at 112kW, and averages 56kW. Dividing by disc area we have:

Peak heat input per unit area = 1.1MW/m^2
Average heat input per unit area = 560kW/m^2

Comparing this with Ben's heat loss figure (earlier post) of ~14kW/m^2, which we might push up to 20kW/m^2 by adding radiative heat loss, we have during braking a peak of FIFTY TIMES as much heat inflow compared with outflow, or 25x on average. The disparity is even greater at lower discs temps because the heat outflow is proportionally less.

This, IMO, is strong evidence for a very UNsteady situation (ie. a reworked version of the Biot number gives ~Bi>1).

6. Two other numbers of significance are the "Diffusivities" of the two materials. Here Aluminium has 20x greater "D" (often "alpha") compared with Stainless Steel.

"D" is the proportionality constant in the Unsteady Heat Equation, so the higher the D, the faster the heat flows away from the surface (the source of the heat during braking) and into the core of the disc.

Conversely, the lower the D, the longer the heat energy stays near the surface, raising its temperature, and increasing the rate of convective and radiative heat losses from the disc.

7. So, IMO http://fsae.com/groupee_common/emoticons/icon_smile.gif, the above seems to confirm what I was getting at in my earlier post. And the important point there was that the Aluminium discs would "turn into jelly" well before any quasi-steadiness was reached.

Now, in order to put some numbers to all this unsteadiness I have to figure out an easy way to integrate the unsteady Laplacian from both sides of the disc (memory http://fsae.com/groupee_common/emoticons/icon_confused.gif ), time permitting!

Z

Ben W

03-08-2012, 05:11 PM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Comparing this with Ben's heat loss figure (earlier post) of ~14kW/m^2, which we might push up to 20kW/m^2 by adding radiative heat loss, we have during braking a peak of FIFTY TIMES as much heat inflow compared with outflow, or 25x on average. The disparity is even greater at lower discs temps because the heat outflow is proportionally less. </div></BLOCKQUOTE>

So what you're saying is there are two different Biot numbers for a disc. One for if it's at a uniform elevated temperature and the surface is then cooled (my use of the Biot number), and the other number is making the convection coefficient the heat input into the brakes. In other words, the rotor pads act as if they are convecting heat into the rotor.

So for cast iron,

h = 1,100,000 W/m^2-k
Lc = 0.00121m
k = 15 W/m-k

Bi(Fe) = 1,100,000 * 0.00121 / 15
Bi(Fe) = 88.73

For the aluminum discs:

Bi(Al) = 1,100,000 * 0.00363 / 200
Bi(Al) = 20.0

I stand corrected. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Buckingham

03-08-2012, 11:56 PM

Let's look at the quasi-steady-state to gather further insight. The car has been running numerous laps and we have recorded on our data logger rotor temperature to be 350 K above ambient (on average, peaks 400 K, valley 325 K).

Let's start with identical steel rotors at all four corners. Take the 56kW average braking power, A=0.1 m^2.

Aware that you can't just use the averages to get the exact answer, but it gives us the ballpark. Required braking power (kinetic energy) is higher at higher speeds, U is probably higher at higher speeds, Qout is higher at higher dT.

Qin = Qout = U*A*dT: 56000 = U*0.1*350
U = 1600 W/m^2-K.
(Note: if you divide Q by 4 rotors and area by 4 rotors you get the same 'U')

U is the Overall Heat Transfer Coefficient and includes both radiation and convection.

Lets say we choose supertonium rotors that due to material properties radiate and convect 25% better.

U = (1600*1.25) = 2000 W/m^2-K
Q = U*A*dT: 56000 = 2000*0.1*dT
dT = 280 deg K

So the supertonium rotors operate about 70 deg K cooler (on average) than the steel rotors.

1.U is large and it almost certainly increases with rotor rpm (possibly close to the same proportion of v^2 that the kinetic energy increases)

2.the hotter the brakes get, the more power they dissipate.

3.Because of (1) and (2), it is difficult to create 'thermal runaway' where the rotor continues to heat up forever. They simply heat up to an average mean operating temperature. If U is lower, average dT is proportionally higher.

4. (1) and (2) is also what keeps the peaks and valleys of dT from straying too far from the average dT.

Regarding core temp vs surface temp:

5. Heat is both generated and dissipated at the surface. Heat flows from hot to cold. Let's say pad area = 20% of rotor area. That means 20% of the time heat is flowing from the pad surface (hot) to the rotor core (less hot). 80% of the time heat is flowing from the core (less hot) back to the uncovered rotor surface (even less hot).

6. If the core was >> hotter than the surface, wouldn't the heat generated at the pad would move laterally along the cooler surface of the pad rather than into the core (path of least resistance!)

7. Because of (5) and (6) in the quasi-steady state system behavior of multiple repeated laps, the net (over a full lap) heat flowing from the surface of the rotor into the core is zero.

Is there a localized transient temperature drop between the surface and the core? Sure.

Does the core/surface reach a quasi-steady state equilibrium such that the net power is zero? Yes

Both aluminum and steel will heat up to the point of thermal equilibrium. If the thermal equilibrium point yields peak dT's too close to the "jelly" temperature of either material bad things will happen.

The issue is that steel can withstand about 2x the dT that aluminum can and aluminum does not have 2x the 'U' for identical geometry as steel.

My point is that the answer to "STEEL OR ALUMINUM?" is actually based on U/dT_jelly and not due to internal conduction properties or mass (notice this is my first mention of mass).

Regarding doing an analysis of a single stop from max speed to zero and sitting there...
what happens to an F1 car if they were to do a decel from max speed to zero and then just sit there at the end of the straight indefinitely??? things melt.

03-15-2012, 06:21 PM

Buckingham,

I agree with most of your above post, except for the following few points. http://fsae.com/groupee_common/emoticons/icon_smile.gif
~~~o0o~~~

You say, "Qin = Qout = U*A*dT..."

I agree that heat energy into the discs/rotors is, on average, equal to heat energy dissipated by radiation and convection. But "Q" is energy per time (J/s, or Watts).

Qin only happens while the brakes are applied, whereas Qout is going on all the time. If the car could accelerate as hard as it brakes (ie. Acc = Dec = 1.6G) and it did only that repetitively, then Qin would be twice Qout (because Qin only occurs for half the time). More realistically, the car spends time coasting through corners and cruising down straights (all the while Qout-ing), so probably Qin >> 5x Qout.

So I'd say U << 300 W/m^2.K.
~~~o0o~~~

Regarding your point 5. I assume this is during braking, when the rotor surface is repeatedly heated by pad-friction then exposed to cooling air.

Here I reckon the core temperature keeps rising (never drops) because the Qin pulses are much bigger than the Qouts, and diffusivity smooths the oscillating surface temperature profile a short distance into the rotor, giving a constantly (though not smoothly) rising core temperature. Easier to show with a sketch, but...
~~~o0o~~~

"... steel can withstand about 2x the dT that aluminum can... "

My quick search on the web suggests that most aluminium alloys are OK to about 250degC, but NO GOOD after 350degC. So "2x" is about right, or maybe even a bit conservative (could be 3x).
~~~o0o~~~

"My point is that the answer to "STEEL OR ALUMINUM?" is actually based on U/dT_jelly and not due to internal conduction properties or mass ..."

This is our main point of difference. I am a firm believer in the philosophy of the "I Ching", ie. "the only constant is change...", etc. I see all phenomena as being transient, although sometimes the transient-ness gets a bit boring, and "quasi-steady".

Consider the family wagon. Drive to the shops and the brake temps fluctuate about a quasi-steady warmish-ness. Even spirited driving on a winding country road only pushes the fluctuations up to around a mild-hotish-ness.

But the real test of brakes comes on "holiday-hill". A fully loaded wagon with equally (over-) loaded un-braked trailer, and a very long, steep, downhill stretch, with a few hairpins thrown in. No quasi-steadiness here! The heat keeps building up in the brakes until one of two things happen:
1. You reach the bottom of the hill and can speed up to get rid of that horrible smell.
2. The brakes melt and you go over the cliff and die.

My point is that with steel discs you get to live http://fsae.com/groupee_common/emoticons/icon_smile.gif, but with aluminium you die http://fsae.com/groupee_common/emoticons/icon_frown.gif.
~~~o0o~~~

The Numbers.
==========
Below is a simple approach to transient heat calcs for discs of arbitrary size and material properties.

I hoped to do this in more detail, namely with repeated and varying Qins and Qouts. However, I only have a simple (+-*/) calculator available (couldn't be bothered using whatever must be in this box, but is cunningly hidden away!) so I am taking some short cuts.

The main condition being modelled is a constant braking heat input, Qin, from both sides of a solid disc. This is the nett of frictional heating inwards, less (the much smaller) radiative and convective dissipation outwards. The main outputs are dTc and dTs, the core and surface temperature rises of the discs.
~~~o0o~~~

The gist of this model is that as the heat flows in from each side of the disc it adopts a U-shaped temperature profile. This is (probably?) best modelled as a "cosh" function (ie. hyperbolic cos, or exponential from each side), but since cosh does not differ much from a parabola, I use a simpler parabolic profile, Temp = F.x^2 + G, with x being distance through the disc (x=0 at core/centreline), and F, G scaling factors to be found from the BCs.

The heat flux at the surface, Qin, determines the slope dT/dx of the parabola at the surface. So F is found by,

dT/dx = Q/(Area.Konductivity) (ie. heat flow eqn.), which = 2.F.L (ie. slope of parabola at x = L = surface), or
F = Q/(2.L.Area.K).

The energy into the brakes, dE = Q x time, together with the material properties of the disc, determines the area under the temperature profile curve. So G is found by,

dE = Q.dT = Area.Rho.SpecHeatC.(Integral of TempProfile.dx (=(F.L^3)/3 + G.L)), or
G = Q.dT/(Mass.C) - (F.L^2)/3.

So,
1. F.L^2 is the deltaT between core and surface (with L = half disc thickness (m)), and
2. G is the deltaT of the core (ie. core temperature above ambient).

It is worth noting that for a given Qin and type of disc, the parabolic heat profile stays the same shape once the heat reaches the core, and it rises steadily up the T axis as more heat energy enters the disc.
~~~o0o~~~

So at last we get to the numbers.... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Let's assume Qin = a constant 80kW for 1 second, which is pretty close to original question of a FSAE car braking from 100kph. The disc sets are as in my previous post (ie. total Mass = 1.6kg, total Area = 0.1m^2, and material properties as before), plus two more sets for comparison.

At the end of this Qin braking phase we have:

Type of Disc ------------------------ dTcore (degC) --- dTsurface (C) ---- dTs-c (C) ---- dTs/dx (C/mm).
================================================== ===============================================
1. Steel
(4mm thick, as before) ------------------ +91 ------------ +144 -------------- +53 -------- 53

2. Hi-K Steel, (Brass?) as above but
with K = 200 (same as Al) -------------- +107 ---------- +111 -------------- +4 ----------- 4

3. Aluminium
(12mm thick, as before) ------------------ +52 ----------- +64 --------------- +12 -------- 4

4. Half-Mass Aluminium!
(only 6mm thick, for "optimality") ------ +109 --------- +115 -------------- +6 ----------- 4
~~~o0o~~~

Conclusions from above:

1. Only a simple model(!), but for given conditions (braking with constant Qin) the temperatures should be reasonably accurate.

2. The regular Steel disc (#1) has the greatest dT between core and surface (53 degC), due to its very low conductivity (note this is Stainless Steel with very low K = 15). It therefore has the hottest surface at all times, and will thus dissipate heat energy sooner and faster than the other discs.

3. The Hi-K Steel/Brass discs (#2), and the two Aluminium discs (#3&4), have almost uniform temperatures from surface to core, due to their high conductivity (K = 200).

4. The optimalised Half-Mass Aluminium discs (#4) will be up to 300+C, and turn to jelly, after 3 or 4 hard stops.

5. Comparing the equal mass regular Steel (#1) and Aluminium (#3) discs on a car going down holiday-hill, (extra calcs required here, but...) IMO the Aluminium will reach jelly temperature long before the Steel, despite its double heat capacity C, and everyone in the car will die.

Now that is a bummer of a transient! http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

(Edit: Added temperature gradient at the surface (ie. dTs/dx = Qin/Area.K) to the above table, in degrees C per millimetre.)

Gruntguru

03-19-2012, 05:32 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Ben W: <BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> It is also clear that the sharply elevated temperature is confined to the surface of the disc (the glow diminishes too rapidly to be more than surface deep). </div></BLOCKQUOTE>

How quickly something cools down is not a direct indication of interior temperatures. Let's assume that you are correct, that the thermal energy pretty much stays at the surface. An otherwise identical rotor, except with a lower heat capacitance and higher conductivity will look the same. It will glow red hot and cool down very quickly. But it will have a hot core as compared to the other rotor.

You can't just look at the surface temperatures to determine the core temps. You can calculate the core temps if you know the thickness, thermal resistance, and heat capacity. </div></BLOCKQUOTE>
There are some problems with your hypothetical material with lower thermal capacity and higher conductivity.
1. Higher speeds (with KE proportional to V squared) would produce much higher temperatures than the cool-core model. (The cool-core model works partly because small increases in surface temp increase surface heat dissipation via all three modes Radiaton, convection and conduction (to the core)). Observation supports the cool-core model over the hypothetical material.

2. If the observer knows he is looking at cast iron or steel and he/she is an engineer, he will have seen those materials heated through to red-hot and air cooled from that temperature. Observation of dynamically cycled brake rotors does not suggest other than the cool-core model. Incidentally I have also seen overheated brake rotors glowing and the effect is totally different - the glow persists for minutes - even in the absence of brake application.

3. The hot core model would require a material with a dramatically reduced thermal capacity for the energy available to produce red-hot temperature throughout. Easy calc.