[jpr]

I recently came across this SAE technical paper co-authored by two brake engineers at GM - The Effect of Rotor Crossdrilling on Brake Performance http://www.sae.org/technical/papers/2006-01-0691

Basically, it looks like they didn't necessarily set out to do a full performance comparison, but incidental to their testing on three brake systems they found they had enough data to write it up as a paper.

It'll cost you $14 do download a copy, but it's well worth it if you are interested in the subject. Hopefully at least some of you are willing to give it a read and start a discussion about it. Below are my thoughts to kick it off.

First off, the surprising conclusion is that drilling the rotors actually does increase the rate of cooling for a given a rotor, by up to about 20% in one case. Unfortunately however, that gain comes with some very serious drawbacks, including the possibility that your rotors may actually run hotter despite of it.

I believe one of the most significant considerations in interpreting the paper's data is that they used semi-metallic brake pads for all the testing. From what I've read, these typically have a max operating temp rating of about 500~550 deg C. Secondly, I understand that the primary functional mechanism of this type of pad is abrasion rather than adhesion. By contrast, the higher temp rated race type pads are more biased towards an adhesive mechanism relying upon a transfer layer of material deposited on the rotor.

Where this comes in to play is in interpreting the test data that shows higher apparent friction and deceleration gain for the plain rotor at low temp, the cross-drilled rotor at higher temp, and generally greater temperature fade stability for the cross-drilled. At low temp, the greater surface area of the plain rotor is the dominant performance factor, as one would expect. At higher temps however, the breakdown of the pad is the dominant consideration and the cheese grater effect of the cross-drilled rotors helps to compensate for this by providing increased mechanical friction.

If however, this test were to be repeated with a higher temp rated adhesive type race pad, I'd expect the results to be exactly the opposite. That is at lower temps where you would expect the race pad to perform poorly, the cross-drilled rotors would provide an abrasive mechanism and increased performance. But at higher temps, not only would the lesser surface area of the cross-drilled rotor become the dominant consideration, but the abrasive mechanism of the holes would inhibit proper formation of the transfer layer.

Another interesting aspect of the paper is the radical base performance difference between the three brake systems used in the test. "System 3" appears to be an overall poor performer in comparison with the other 2 systems. In the case of System 3, the cross drilling appears to be a pure bling change, as it is rather consistently outperformed by the plain rotors, despite their being smaller in overall diameter. System 1 was a better performer, but quite unbalanced. Almost all the work on this system was being done by the front rotors. System 2 looks to be the best of the lot, and consequently showed the least radical performance deviations between the plain and cross-drilled rotors.

I believe the base performance difference between System 1 and System 2 helps in interpreting the results showing increased cooling efficiency with the cross-drilled rotors. On System 1, tests at 50kph, 100kph, and 160kph, showed an increase of cooling values (hA) of 8.8%, 12.1%, and 20.1% on the front and -3.2%, 1.9%, and 8.5% on the rear. For System 2, the values at 50kph, 110kph, and 140kph are 7.8%, 10.4%, and 12.1% for the front and 4.1%, 7.7%, and 13.4% rear. The other key piece of data is the raw hA numbers. At 160kph, the average hA for the front rotors of System 1 is 19.26 for the plain rotors and 23.13 for the cross-drilled. By comparison, the System 2 values at 140kph are 23.35 for the plain rotors and 26.18 for the drilled. The conclusion I draw from this is that while cross-drilling does improve the cooling rate, the degree of benefit is inversely proportional to the size of the rotor. From a pure cooling rate consideration, the plain rotors of System 2 outperform the cross-drilled rotors of System 1. In practical terms, using cross-drilled rotors would seem to make the most sense only after you have run out of options for increasing the rotor diameter.

Pad wear on cross-drilled rotors was also shown to be up to up to 50% greater under hard use, and about 25~30% greater under street use. What was not explicitly tested was the effect of this increased wear on temperature. As I understand it, the thinner the pad, the higher the pad temp. My suspicion is that at some point, the diminished heat capacity of the pad due to wear more than offsets the gain in cooling capacity of the cross-drilled rotors. I further suspect that this may be the operating mechanism behind many of the catastrophic failures of cross-drilled rotors at the track. Basically, they will start off strong, but the accelerated rate of pad wear soon drives the rotor into an overheated condition.

Some of the other data and conclusions from the paper on cross-drilled rotors:

• There is no practical difference in wet weather performance.
• Pad outgassing is not factor
• Thermal fatigue life is significantly shorter for cross-drilled rotors. In the case of poorly balanced and possibly under-sized System 1, by as much as 50%.

 

[hvirani]

Excellent information, thanks for posting.

 

[shortyb]

jpr-as usual, some very informative info coming from you.

Didn't download the paper, but have read and talked at length with brake folks about this same topic. Seems in the area of cooling, the space between the rotor plates (vane case) in ventilated rotors contributes the most to convective cooling. The shape of the vanes (curved, straight), number of, and their width can dictate how well a rotor will cool in service. In some high speed applications, the vane case is designed to create a pumping action so heated air moves more quickly. What was discovered in some of these designs was that crossdrilling actually REDUCED the convective action by interrupting the pathway for heat to travel effeciently. Just like in a tuned intake system (for lack of a better analogy) the overall length of the "runner" aids in the evacuation of heat. When multiple pathways for convective heat are added, the speed and flow volume are decreased. Couple this with the reduction of mass (holes) and the subsequent reduction in conductive heat transfer, cross drilled rotors can reach heat saturation much quicker than a well designed plain rotor in high speed apps. Obviously, a well designed cross drilled rotor will take these considerations into account when the rotors are speced for a specific service. But most cross drilleds out there for consumer use are just regular vaned rotors that are drilled in a certain pattern.

In your pad type discussion, you mention the frictional function of semi-metallics. All pads will operate in both the abrasive and adherent friction modes. Semis seem to work best in a consistantly higher heat range such as racing and hard track use. They will basically revert to abrasive mode when cool and thats why they don't seem to be the best first choice for street applications. More and more manufacturers are designing pads that operate primarily in an adherent mode. Adherent friction produces the most effecient braking and pad material dictates where the point of friction codeffecient will be greatest. You can still get a race pad to be run in the adherent mode with cross-drilled rotors. But its effect is tempered by the reduced area for pad film to be transferred to, not so much because of the pads run in the abrasive mode.

 

[jpr]

Thanks for the feedback shortyb - I was hoping you would chime in on this.

Everything I've read confirms you point about the importance of vane design being perhaps the most critical aspect affecting the cooling rate of the rotor. From what I understand, curved vanes are best, with pillar post type coming in next. Bringing up the rear are the basic straight vane designs such as on our OE rotors.

Likewise, I'd seen the reference before to studies where the drilling reduced the cooling capacity by interfering with airflow, not to mention the reduction in mass. Which is what make the SAE paper so interesting.

They compared the cooling capacity of the drilled and the plain rotors for the first two brake systems tested. System 1 used a pillar post front and a striaght vane rear. Tests at 50kph, 100kph, and 160kph, showed an increase of cooling values (hA) of 8.8%, 12.1%, and 20.1% on the front and -3.2%, 1.9%, and 8.5% on the rear. For System 2, the front rotor is a curved vane design and the rear a pillar post type. The test derived cooling capacity values at 50kph, 110kph, and 140kph are 7.8%, 10.4%, and 12.1% for the front and 4.1%, 7.7%, and 13.4% rear.

Another interesting data point is that the difference in working mass between the crossdrilled and plain rotors was never more than 0.2kg for all the tested systems. This is a change of less than 5% in the working mass, which doesn't seem terribly dramatic.

Regarding pad types, I think a lot of it depends upon what one considers high temperature. The semi-metalics do seem to have be considered a high temp pad in that they provide a steady frictional coefficient up in the 350degC to 550degC range but still work well cold. I believe this is different from a true race-type pad that will work well up to 800degC or more, but perform very poorly below 150degC or so.

Here's some of my reference sources on the different behavior of semi-metalic vs race type pads - http://www.powerbrake.co.za/download..._01_judder.pdf

To understand what happens next we need to briefly discuss two broad categories of brake pads. The first category of pads relies more heavily on the concept of abrasion to slow your car down. In simple terms this involves the mechanical gripping and breaking off of opposing pad and disc materials at a molecular level. The materials literally wear each other away in the process. The harder material (typically the disc) will wear slower than the softer friction material of the brake pad.

The second category of pads relies more heavily on the concept of adhesion. In this case some of the pad friction material is transferred to across to the contact surface of the disc, where it forms a thin, uniform layer of friction material. Essentially you now have the friction material of the pad coming into contact with identical friction material that has been deposited on the surface of the disc. Under braking, the bonds between the friction material of the brake pad and the friction material deposited on the disc are constantly breaking and reforming. Material crosses the pad/disc interface in both directions as the molecular bonds continually break and reform.

All pad formulations use a combination of both abrasion and adhesion but some pads (such as common semi-metallic formulations) rely more heavily on abrasion and others (such as many high temperature ‘Ferro-Carbon’ fast-road and race formulations) rely far more heavily on adhesion. Abrasive and adhesive pads affect the formation of DTV in different ways.

and here - http://www.eurac-group.com/technote4.htm

Friction material formulation solutions have led to the establishment of two principal schools of thought. The first is to produce friction materials exhibiting a sufficiently mild rotor wear response such that D.T.V. generation will be alleviated. This is termed the passive approach and was originally associated with Japanese friction material manufacturers. Unfortunately, the disadvantages associated with this solution are that manufactured or pre-existing D.T.V. will likely persist throughout the lifetime of the brake, this being combined with difficulties in attaining sufficiently high dynamic friction coefficients. Several associated shortcomings have been attributed to this formulation methodology, the majority of which focus on the material's reduced capability in effecting rotor clean-up in the event of rotor surface irregularity, i.e. friction instability-derived B.T.V. effects.

The second approach is to produce friction materials exhibiting sufficiently severe rotor wear response that any D.T.V. is removed by an in situ machining, planing or truing of the disc geometry during normal braking. This is termed the aggressive approach and has tended to become associated with German friction material manufacturers. This also helps in the removal of physico-chemical inconsistencies at the brake interfaces such as uneven transfer film deposition or corrosion products. The down-side to this approach is obviously the tangible reduction in rotor life but also that the heavily abrasive action of these materials mean that hot, electrostatically-charged, metallic wear debris being ejected from the friction pair is deposited unattractively on neighbouring ventilated wheel rims. Indeed, the nature of this coarse debris results in the effect being partially permanent on plastic hub caps or plastic-encapsulated, alloy wheel rims. This has become especially significant in areas of the US market where hot and dry weather patterns mean that wheels are infrequently rinsed.

The wear response to both the passive and aggressive friction material formulations described, with respect to increasing operating temperature during a drag test to failure, is shown in figure 15. With a greater understanding of microscopic tribological interactions such as physico-chemical phenomena and their influence on wear behaviour, a third and potentially more elegant friction material solution has emerged in recent years. Studies investigating the effect of low contact forces, such as those associated with O.B.W. and consequently, with D.T.V. generation, have noted widely varying rotor wear response to different varieties of friction material.

 

[shortyb]

I'm so intrigued by the cooling capacity numbers, I might just put out the $14 to see how they are coming up with that info.

Interested to see what the control numbers for hA are and compare the delta to that. It may be less of a difference than what is aluded to on paper. Just like what you are theorizing in your first post regarding the pad wear/heat over time vs. cooling capacity. I think you have hit the nail on the head with that one and it can be summed up as a very miniscule gain if someone is buying C/D rotors to run "cooler". Also would like to know what the composition of all 3 rotor sets is. A minor change in gray iron or an alloying metal can effect conductive/convective cooling and could skew the numbers as well.

My reference with regards to pads was more of a racing appropriate semi-metallic wherein the pads go "abrasive" when operated below their "sweet spot" temp range. Semis on the street are just old school IMHO. Both of your posted source material on pads are spot on. Both of them also refer to the advancements in pad material such as ferro-carbon and (even though not mentioned) ceramic composites. I suppose we can thank the US market for pushing these materials into the consumer realm. After all, who b!tches more about dusty/dirty wheels than us? If it weren't for that, we'd all still be cleaning our wheels everyday and replacing rotors at the same time the pads go south no matter who manufactured them. Still wish I could find a pad that will abrade/microplane the rotor for a fresh surface, provide a high percentage of adherent friction, remain quiet, produce low dust and not eat the finish off of my wheels. Also I'd like massive amounts of stopping power, bite, and modulation, please, whether hot or cold. I'll wake up now and smell the offgassing binder material .

 

[jpr]

h = convective heat transfer coefficient
A = working area of brake rotor
They use this as they state "This provides the best measure for a brake rotor designs overall heat rejection capability, and comparing hA values allows for thermal performance of rotors of dissimilar working masses to be compared."

It's derived from the equation h*A=b*c*M where
M=working mass of rotor
c=specific heat capacity of rotor material
b=cooling coefficient of brake rotor

It's that last one, b, that they derived empirically from the equation
b= (1/t)*ln[(T-TsubS)/(TsubI-TsubS)] where
t=time
T=rotor temperature
TsubI=rotor intial temperature
TsubS=rotor stabilization temperature

The procedure used was to heat the rotors by moderate brake snubs until rotor temp was in the 400~500 degC range. Then they drove at the specified speed and allowed the rotor to cool until a stabilization temperature was reached.

The paper unfortunately does not divulge the details of rotor material composition, but rumor is that "System 1" was a Corvette and "System 2" a Porsche, for whatever that's worth.

One of the really fascinating charts in the paper is Figure 12, where they map the average lining temperature vs. stop number for System 3. In this case, the front brake temperatures pretty much track together until stop #9 and a temp of about 375degC. After that, one of them more or less stabilizes at around 440degC by stop 20 while the other shoots dramatically up to a peak temperature of around 670degC. The one with the dramatic temperature rise was in fact the crossdrilled rotor which was clearly outperformed by the plain rotor despite it's being 4% smaller in diameter. Unfortunately, they did not also perform this test on Systems 1 and 2.

 

[armour767]

damn my liberal arts degree....damn it to hell. :banghead:

 

[jpr]

damn my liberal arts degree....damn it to hell. :banghead:

Oh that was funny - :rofl:

 

[armour767]

Oh that was funny - :rofl:

I'm here all week.....seriously though...great post.

P.S. those cross drilled suckers sure do look pimp.....right? :str8pimpi

 

[turbo1168]

Another great article! Back quite a few years ago, I owned a Dodge Omni GLH turbo. I was forever having problems with brake rotor runout. The car was quick from the factory and I had done some mods to it beyond that. But with a car that could do over 130mph and get there quite quickly, faster to 100 than most Mustangs, Camaros, Corvettes, etc... the brakes were substandard. Being that I had little extra money at the time, upgraded complete brakes were unattainable. Instead, a friend that owned a H2O VW performance shop owned a CNC mill that he had programmed to drill rotors. I don't remember the pattern he used but they looked sweet, topped off with a new set of Raybestos Super Stops, Fresh rebuilt calipers, and DOT4 fluid. After that setup was installed I broke em in proper, and was amazed at how well my car stopped after that. The brakes never again pulsed and it stopped better than it ever had before. My father's wife drove a 92 Caravan, that the front brakes were constantly pulsating. Replaced rotors and pads numerous times, still would always start pulsating after a period of time. Tried different materials of pads. We did the same thing on this van with great results, had a new set of rotors crossdrilled and installed fresh Super Stops. Never a problem again. The Omni was a light vehicle (about 2000lbs. after it got put on a diet) but it saw extreme driving, almost short of race conditions sometimes. The van was a heavy vehicle with a heavy footed driver. Between the last thread about rotor "warping", and this one about cross drilling, I have learned some more.

 

[rovert]

I'm here all week.....seriously though...great post.

P.S. those cross drilled suckers sure do look pimp.....right? :str8pimpi

I know mine do! impin: I just got them hoping that when I'm in a rainstorm which Vancouver has a lot....I can stop a bit quicker right away when i jam on the brakes while the rotors are soaking wet and cold. My OEM flat pads had no friction until half a second after I went on the brakes hard which is scary when I can't slow down because there is water in between the rotor and pad. :banghead:

 

[flyinj3x]

how about this..

if airflow through the drilled holes is a factor, wouldnt the brake pad close those holes when they passed under it? and when the holes are closed, the cooling vanes are still trying to pull air from these locations? wouldnt there be a slight vacuum created on the pad as it passes over the holes? if so, would this have an effect on adhesion? yes the holes are reduced area for the pad to adhere to, or wear on. would this small vacuum have a positive side? it actually helps the pad bite. idk, thought id run it by the experts..

 

[jpr]

Good questions, I'll try to fill in some of the missing pieces of information..

First off, no significant cooling effects take place during braking. While the brake is applied, the two dominant considerations are how fat you are putting heat in to the rotor and how fast that heat is getting dispersed within the rotor. This is where rotor mass makes a big difference as it gives you more room to absorb more heat.

Airflow through the rotor is from the center towards the edge. Since the edge of the rotor is moving faster than center it creates an air pressure differential, which is what makes the air move. Curved vane rotors obviously help this by creating an additional pumping acation, but the air pressure differential explains why air flows through straight vane and pillar post tyoe rotors as well.

So when it comes to the holes, what they are really doing is creating an alternate exit path for the air before it gets to the edge of the rotor. The authors of the paper thought that this may explain some of the test results on System 1 where at low speed the drilled rear rotors actually ran hotter than the plain ones. They surmised that at low speeds there was not sufficient force generated to draw air all the way out to edge of the rotor. They do note however that this result may be peculiar to the design of these rotors, which featured a "very short airflow path" and a relaatively large number of crossdrill holes versus the brake plate area.

 

[shortyb]

how about this..

if airflow through the drilled holes is a factor, wouldnt the brake pad close those holes when they passed under it? and when the holes are closed, the cooling vanes are still trying to pull air from these locations? wouldnt there be a slight vacuum created on the pad as it passes over the holes? if so, would this have an effect on adhesion? yes the holes are reduced area for the pad to adhere to, or wear on. would this small vacuum have a positive side? it actually helps the pad bite. idk, thought id run it by the experts..

There really isn't a vacuum per se, It is the pressure differential that results from the heating, ie; hot air rises etc. and the speed at which the rotor travels. Basically it is convection that causes the cooling "pathway" through the vane case. This phenomenon is greater as speeds increase due to as jpr has said, the speed/pressure differences at the edge vs. the hat.

Mass plays a much more important part in the overall cooling equation. This is why rotors on track machines have very thick plates, seperated by a wide vane case, and are as large a diameter as can be stuffed into the wheel barrel. Cross-drilling will reduce this mass, but it seems the primary reason for racing is that it saves weight. Kinda strange to think that if mass aids in cooling, why would they reduce it? It comes in the form of balance. Designers can spec a very large, massive rotor that will absorb/shed heat at "x" rate under severe service. If x exceeds the threshold of heat generation (component of pad compound, rotor interface etc.) then they can reduce overall mass (cross drill) and stay within the preferred heat range. So why not just use a smaller rotor? Swept area and brake torque. Having a longer "thinner" pad nearer the rotor edge will produce more torque. This also allows the rotor sufaces to be smaller and the hat to be made of lighter materials. It can also be physically bigger in diameter and "float" the rotor on pins or buttons. Doing these small things reduces two of the killers of speed and direction change: inertia and momentum. Just like we all learned in physics class (liberal arts degree or not) is that objects like to stay in motion and on a straight path. It takes force to overcome this and in simple terms, it takes a smaller bully to take a smaller kid's milk money. Bigger Poindexter? Bigger Crusher. Or more little Crushers.

This all makes sense in racing but not so much for regular street use. And since adherent friction is a much better friend than abrasive, you should give it as much room to play as you can. Street use brake heat is well below what the components can handle. Manufacturers build in a certain amount of safety margin in the generally unlikely event you need it (braking down a long grade, running from the cops, etc). So buying a rotor that is cross-drilled for the pure intent of "running cooler" is not necessary. For "looking cooler" it may be a good investment, but spend you money on better tires or pads if you want better stopping power.

jpr-wish they used same rotors plain, then cross-drilled them and retested. But then we would get into the gray areas of tempering and heat risers etc. Oh well. Just like with getting to the center of a Tootsie Pop, one may never know .

 

[mkodama]

:bump: for a lot of good info

Thank you jpr for sharing that $14 article with us. :luv: