Brake Aero Cooling

[olefud]

I've been working on brake cooling using aero means to reject heat before it flows into the rotor. So far an artistic success.

More specifically, I use a vane closely spaced adjacent each rotor face to remove the boundary layer. This, as it turns out, yields significant rotor cooling but raises the pad temp.

My thinking was that the cooler rotor would also cool the pad by conductive heat flow. The boundary layer was thought to be too hot to be useful. However, it appears that the very hot boundary layer may be carrying cooler air with it that may provide important convective pad cooling.

Anybody looked at aero characteristics of boundary layer relative to cooling?

[riff_raff]

"Anybody looked at aero characteristics of boundary layer relative to cooling?"

olefud,

With regards to your brake cooling question, the thin boundary layer is where the heat transfer actually occurs. Two of the variables in heat transfer are the relative temperature difference between the airflow and rotor surface, and air mass flow. The relatively cool, thin boundary airflow has very little mass and thus heats up very quickly. You somewhat alluded to this issue in your post.

Manufacturers of air/air or liquid/air heat exchangers have long known that tripping up the boundary airflow within the core passages, and making it turbulent, improves the heat transfer process.

So you don't actually want to "remove" or inhibit your rotor cooling airflow. What you really want to do is to create a flow condition where the boundary and core flows constantly turn over, or tumble, as they pass over the hot rotor surface. You did not provide details of your concept, but since the boundary layer is very thin, even something like a thin wire(s) located very close and parallel to the rotor surface, and transverse to the airflow, would do the trick.

riff_raff

[olefud]

Thanks riff raff. I think I see something different regarding the rotor. First, the laminar flow boundry layer is rather insulating. Second, since air, rather counterintuitively, becomes more viscous with temperature, the boundary layer becomes a thick, insulating layer on the rotor. And it is adhered so strongly that directed convection cooling flow does not displace it.

I have dyno data that confirms this theory. With vanes I get substantially lower rotor temps relative to a control. The surprise is that the pad temps go ballistic. Originally, I thought that conduction cooling of the pad by the rotor would more than offset any loss of convection cooling by the hot boundary layer. My thought now is that the rather thick rotor boundary layer also drags along a good bit of cool air with the very hot that carries away pad heat. In any event, at this stage, I have a nicely cool rotor and a smoking pad.

I have a long shot hope that someone might have looked at the boundary layer/pad interaction. The problem can be solved but I don't want to reinvent the wheel if there avilable data out there.

[riff_raff]

olefud,

Your comment about the attached boundary layer airflow being "insulating" is essentially correct. Air is actually quite a good insulator.

Think of your heat transfer situation in simplified terms. First, you have convective heat transfer from the hot metal rotor surface to the thin boundary airflow. This thin boundary airflow is initially cool, but quickly heats up when it contacts the rotor surface, due to the large temperature delta and low mass of the boundary airflow. Initially the heat transfer rate between the rotor surface and cool boundary airflow is high, but as the boundary airflow temperature increases the heat transfer rate slows, until it finally stops when temperature equilibrium is reached between the two. Second, you must consider the heat transfer rate between the boundary and core airflows. Since the specific heat value of air is lower than metal, the heat transfer rate at this interface is much lower than that at the boundary flow/rotor surface. This greatly simplified explanation assumes, of course, that the boundary airflow is truly laminar in nature, and there is no mixing between the boundary and core flows.

With regards to brake pad temps and airflows, most friction materials have very low thermal conductivity through the thickness. With something like conventional organic fiber pads and iron rotors, the friction heat load generated at the contact is effectively transferred out through the large mass/large surface contact area/high thermal conductivity present in the iron rotor, but the small mass/small surface contact area/low thermal conductivity present in the pad results in the pad surface becoming very hot. Since there would be no airflow occurring at the intimate pad/rotor contact during braking, the whole issue of air cooling here is somewhat irrelevant.

High performance brake systems take two different approaches to this basic problem. One is to make the pad material itself more thermally conductive, which is what semi-metallic pad materials do. The other is to use pad/rotor materials that can tolerate extremely high temperatures, which is what carbon/carbon brakes do.

riff_raff

[olefud]

Hey riff_raff, I sort of feel I've been sandbagging you a bit. This is a project I've been working on for a while that, I belive, is actually cutting edge. You've been good enough to work up a pretty good tutorial on brakes for me. I'll try to reciprecate. I'm under a confidential agreement so I can't tell all. But I did enough IR&D that I can outline the basics of what I'm doing.

It all started when I was road racing a heavy, underbraked sedan (solid rotors and drums). Following pretty much what you suggested I got fairly good brakes. But no matter how much free air was ducted to the rotors they heat checked and warped. I suspected the boundary layer was insulating againt the cooling flow. From some other work I was aware that the air viscosity incrases with temp. Checking the Navier-Stokes equation, the viscosity term convinced me that the boundary layer was in fact rather thick.

Using a powered axel to drive the brakes while exposed on a rack, I noted that the tell-tale incandescent particles also appeared to be in a thick, stable layer. The same particles showed that an aero vane would detach and direct the boundary layer.

A bit of background that you already know. When used hard, brakes reject heat primarily through radiant cooling which increases at something like the fourth power of the absolute temperature. Short term, the rotor serves more as a heat sink while convection rejects heat at a slow but steady rate. The problem is that to bring radiant cooling into full play, the rotor must operate close to its failure temp.

Thus the importance of forced convection with the aero boundary layer vanes(to be clear, the vanes are adjacent the rotor swept areas) is that it is effective at lower temps than radiant cooling, and with increasing temps becomes even more effective as the boundary layer thickens. The aero vane mechanism is also additive to the other heat rejection means. Dyno runs show the rotor doesn't reach high temps even under severe tests.

I didn't want to get into this too deep since my problem is cooling the pad. This doesn't appear to be too much of a problem -at least when ducted blown or slipstream air is available.

My original inquiry was more to the boundary layer/pad cooling dynamic. If data on this exists, it's probably proprietary. Your efforts are appreciated.

[marcush.]

a collegue of mine did cool the backplates of the pads with liquid (The backplates were crossdrilled and a coolingflow was introduced to reduce pad temperature .worked very well.

[olefud]

a collegue of mine did cool the backplates of the pads with liquid (The backplates were crossdrilled and a coolingflow was introduced to reduce pad temperature .worked very well.

The heat content of pads is much less than that of the rotor. Thus cooling should not be too troublesome. Your suggestion would be effective if competition rules allow.

[riff_raff]

olefud,

I guess I don't understand what you're describing.

Basically, an automotive disc brake is a device that converts the vehicle's kinetic energy into heat, via friction at the pad/rotor sliding contact. This friction heat is primarily transferred away from the pad/rotor contact interface via conduction in the pad and rotor structures, and not thru convection to the airflow. Since the pad is in intimate contact with the rotor surface when the heat load is being generated, there is very little potential for convective pad cooling at this point.

The conditions that exist at the pad/rotor interface, with the small area of the pad surface and the pad's low thermal conductivity, combined with the large area of the rotor's working surface and the rotor material's higher thermal conductivity, means that most of the conductive heat transfer takes place within the rotor, and not the pad.

With disc brakes, even after the brake caliper pressure is released, the clearance between the pad and rotor surface is still extremely small. And thus there is very little potential for cooling airflow thru this space. Most high performance disc brake calipers use extensive measures to thermally isolate the pad back plates and pistons. The piston/pad contact area is minimized, and piston materials with low thermal conductivity (such as titanium) are used.

If you have an overheating problem with your brake pads, the logical approach would be to:

A) Increase your pad contact surface area (ie. 6 piston calipers, or dual calipers per wheel). This would also necessitate higher caliper clamping forces.

B) Increase the conductive heat transfer rate away from the pad/rotor contact. This would require higher thermal conductivity in your pad and/or rotor materials.

C) Use rotor/pad materials with a higher thermal strength limit, such as CRC.

Finally, if you want to see an extreme example of how disc brake systems are designed to handle massive braking energies, take a look at a commercial aircraft disc brake system. Lots of rotor surfaces and lots of pad surface on each wheel.

http://sitelife.aviationweek.com/ver1.0 ... .Large.jpg

[olefud]

Since the pad is in intimate contact with the rotor surface when the heat load is being generated, there is very little potential for convective pad cooling at this point.

So I thought too! What I found was a new and cumlative heat rejection path using aero to remove the very hot boundary layer from the rotor. The layer is very hot because, being laminar, it is insulating. What I'm doing is peeling this layer and the heat energy therein from the rotor and allowing cool air to reach the rotor. I don't imagine the heat content is that great each event, but it happens at least once each rotation so it area under the curve is substantial. Brake dyno testing has shown greatly reduced rotor temps under extreme conditions.

My expectation was that disruption of the heated boundary layer would have little, or even a favorable, effect on pad temps for the reasons you state. However empirical data show the pad temps in fact increase when the boundary layer is disrupted. This isn't an overwhelming problem since the heat content of the pad is rather low. But I would prefer to address it in the least involved manner. That's why I was asking about insights into boundary layer/pad interaction.

My view was in accord with your description, but I can't argue with the data.

[phils378]

Hi Olefud
Just read the threads, did you make any progress with boundary layer cooling? Did you ever investigate if grooves in the faces of the rotor, like you see on most iron race discs, achieve the same thing as vanes close to the surfaces?

[olefud]

Hi Olefud
Just read the threads, did you make any progress with boundary layer cooling? Did you ever investigate if grooves in the faces of the rotor, like you see on most iron race discs, achieve the same thing as vanes close to the surfaces?

Check out the current Brake Heat Rejection Concept thread that’s a bit more comprehensive. The grooves and x-drilled holes in a rotor are essentially to vent gas pressure under the pad. If anything, they would promote a thicker boundary layer by increasing surface roughness, but I doubt that there’s much boundary layer effect.

[riff_raff]

Interesting topic. After thinking about it for a while, here's an approach that might be worth considering. What about cutting helical grooves into the pad that start at the pad's leading edge and move radially outward. These grooves would allow some of the boundary airflow to continue flowing over the pad surfaces providing significant pad cooling. The configuration of the pad grooves would also produce even wear rates across the pad/rotor interface. The only change that may be required with carbon/carbon brakes is an increase in the surface area of the pad to compensate for the cooling grooves, in regards to pad wear rates.

[olefud]

Interesting topic. After thinking about it for a while, here's an approach that might be worth considering. What about cutting helical grooves into the pad that start at the pad's leading edge and move radially outward. These grooves would allow some of the boundary airflow to continue flowing over the pad surfaces providing significant pad cooling. The configuration of the pad grooves would also produce even wear rates across the pad/rotor interface. The only change that may be required with carbon/carbon brakes is an increase in the surface area of the pad to compensate for the cooling grooves, in regards to pad wear rates.

Interesting thought. I did something like this with grooves and holes in the pad. However, my thought was to relieve the gases generated by pad sans x-drilling or grooves in the rotor.

I don’t know exactly what happens at the leading edge of the pad/rotor interface; but the net result is a hotter rotor and cooler pad in the conventional arrangement. I’d like to see some IR pictures, with and without the vanes.

[cwb]

Hi Olefud,

I dont think that the effect you are seeing has anything to do with air getting more viscous at high temperatures, or boundary layers getting too thick. I would focus on the bulk parameters of convective heat transfer which are the mass flowrate (function of velocity) and the bulk temperature difference between the item being cooled and the airstream.

If the rotor is running cooler when you introduce your vanes then I suggest you have either increased the velocity of the airstream over the rotor, or more likely, your vane deflects away air that has already been heated by the rotor, and this air is replaced with cooler air, thus you are increasing the bulk temperature difference between the rotor and the airstream over the total surface of the rotor and increasing the amount of heat removed.

Similarly with the pads, if they are running hotter then I would guess your vanes are somehow reducing the bulk airflow over the pad/carrier assembly and reducing the cooling effect. As you rightly say because of the proximity of the rotor there probably isnt much of a convective effect, likewise because it is a 'floating' element it doesnt have much physical contact to conduct heat away, so it wouldnt take much of a change in conditions to cause an increase in temperature.

[olefud]

Hi Olefud,

I dont think that the effect you are seeing has anything to do with air getting more viscous at high temperatures, or boundary layers getting too thick. I would focus on the bulk parameters of convective heat transfer which are the mass flowrate (function of velocity) and the bulk temperature difference between the item being cooled and the airstream.

If the rotor is running cooler when you introduce your vanes then I suggest you have either increased the velocity of the airstream over the rotor, or more likely, your vane deflects away air that has already been heated by the rotor, and this air is replaced with cooler air, thus you are increasing the bulk temperature difference between the rotor and the airstream over the total surface of the rotor and increasing the amount of heat removed.

Similarly with the pads, if they are running hotter then I would guess your vanes are somehow reducing the bulk airflow over the pad/carrier assembly and reducing the cooling effect. As you rightly say because of the proximity of the rotor there probably isnt much of a convective effect, likewise because it is a 'floating' element it doesnt have much physical contact to conduct heat away, so it wouldnt take much of a change in conditions to cause an increase in temperature.

We’re on the same page except for the viscosity thing. The point is that air at high temperature becomes more viscose which, in turn, forms a thicker, more adherent, insulating boundary layer. The boundary layer, being laminar, transfers heat by simple conduction that is rather ineffective, particularly during the short cycle period between the trailing and leading edges of the pads during which the transfer occurs. This is all rather counterintuitive in that the physical properties of air change appreciable at 800° to 1000° F. Checkout the Jan 30 discussion in the related Brake Heat Rejection thread.

As you recognize, the vane essentially “pumps” the heated boundary layer from the rotor. That’s the purpose of the vane, to take advantage of the thick, hot boundary layer to reject heat.

It was a bit of a surprise to me that the pads run hotter with the vane. I had expected the cooler rotor to conduct heat from the pads. But it appears that the boundary layer also carries cooler air to the pad that is also disrupted by the vanes. Since the pads are not very heat conductive and have a fairly low specific heat, this development is hopefully not that much of a setback.