Thermal management on roots type supercharger 3

[yoshimitsuspeed]

Many of my customers run 4AGZE engines that run an old Ogura two lobe roots supercharger.

Many people are running these well outside their original design parameters. At a certain point the lobes hit the housing and it wrecks the supercharger. Whenever I can I try to talk people into going a different direction and use something designed for higher pressure ratios but sometimes people just want to get as much as they can out of the stock unit.
Most discussion is usually focused on cooling the compressor housing but I started wondering if cooling the outside may actually do more harm than good. The temp of the housing will have very little effect on the temp of the lobes. On the other hand if the lobes heat up and expand a lot and the housing is cool and expands less it will be more likely for the lobes to hit each other or the housing. This has me wondering if it may actually be better to coat the housing with a ceramic thermal barrier which would help keep the housing temp more even and closer to the temp of the lobes. If everything heats and expands a similar amount then the gap between the lobes themselves and the lobes and housing should stay closer to nominal shouldn't it?
I know it would be more ideal to work out a way to cool the lobes but that's not going to be a viable option.

 

[Tmoose]

In the old days dragsters adapted GMC blowers/superchargers and soon were overdriving them.

http://cacklefest.com/images/Picture-151.jpg

I seem to recall magazine articles about modifying the rotors for desired clearances, teflon wear strips, and hard coatings before they evolved to manufacturing specialty rotors and then complete superchargers.

 

[Compositepro]

I agree with your reasoning that cooling the casing might be a bad idea. Plus, it will not be very effective for cooling the compressed air.

 

[1gibson]

SMALL amount of water/methanol injection before the blower should keep temps down, and increase volumetric efficiency.

 

[winfieldblue]

I believe that these Toyota chargers are not rebuildable, so coating the rotors is going to be difficult!
As Gibson said, water/meth injection could be the best way to go.

 

[yoshimitsuspeed]

They are not rebuildable but you can get to the rotors. You are right that they aren't rebuildable. Actually that's not even true. There are no bearings available in the aftermarket that will fit the unique size they used on the lobe bearings. With custom bearings or customizing the housing for different bearings it is fully rebuildable. It just puts it out of the range of practicality for an old inefficient roots blower.
Above I wasn't talking about coating the rotors though. I was talking about coating the SC housing to try to keep heat in and more evenly distributed. The only downside is that then you are introducing more heat into the oil and bearings but I don't imagine the temps should be drastically different.
Really I am just wondering what I should coat it with. I do plan on ceramic coating it to make it look nice but I have three options. I could use heat dissipation coating, Thermal barrier coating or a regular coating that wouldn't have a significant impact on heat dissipation. Just trying to decide which would be best.

 

[Lou Scannon]

The exit air temperature will go up exponentially with the pressure ratio, and I suspect the internal and casing temperatures will track the exit temperature to some extent.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[140Airpower]

yoshimitsuspeed, IMHO even a thin thermal blanket will be much more effective than a coating, but both together should give you what you want.

 

[Tmoose]

Any way to quantify, at least a little bit, the "certain point the lobes hit the housing and it wrecks the supercharger" ?

 

[yoshimitsuspeed]

140Airpower
I don't plan on doing or suggesting doing more than a coating. I wouldn't be surprised if it made a huge difference one way or the other but I just thought I would ask since as often happens my logic and understanding of things contradicts the standard beliefs and so I always want to try to wrap my head around all possibilities as much as possible. The last thing I want to do is make the SC able to take 10% more heat without tapping the rotors in trade for dying 5 times faster due to overheating bearings. I doubt that's going to happen but these are the things I am trying to weigh out before settling on a decision.

Tmoose
There would be a lot of variables to fill in to come close to predicting this. Ultimately it all comes down to CTE. Theoretically if the whole supercharger heated up evenly it should be able to get incredibly hot without anything hitting. The case is aluminum, the lobe structure is aluminum which is coated with some form of Teflon coating. Since the housing and the lobes are the same material if they heat up evenly and expand evenly then they should maintain near nominal tolerances. This is why I think that when the supercharger blows up due to too much heat that it must be that the housing is not heating up as much or as fast as the lobes. If the lobes heat up and grow more faster then they will get closer to the housing and eventually touch.
You could work this all out on paper pretty accurately but it would take a good bit of work and filling in many variables to get close.

 

[140Airpower]

yoshimitsuspeed, Interesting concern, bearing life. Keep in mind that this thing is air cooled. That may sound funny when the air is hot, but the assembly can't get significantly hotter than the compressed air..., unless the bearings themselves or the rotor/housing contact make a lot of heat and then the airflow through it still puts a limit on the temperature.

 

[Compositepro]

Keep in mind that aluminum loses strength very fast above 300F. Heat treatment temperatures are as low as 350F.

 

[yoshimitsuspeed]

Unfortunately roots blowers are horribly inefficient. Fortunately the housing shouldn't get anywhere near the temp of the air coming out. at 16 PSI boost and 45% adiabatic efficiency on a hot day you could see outlet temps in excess of 400f. Fortunately enough the SC only sees that for a fraction of the time. Half the SC housing is getting a constant flow of ambient air flowing past. The lobe spend half their time exposed to ambient air and half their time the warmer compressed air. This incoming ambient air will keep the SC much cooler than outlet temps. Unfortunately that does mean the outlet side of the compressor will tend to be much hotter than the inlet side.
I could get real fancy and coat the inlet half of the housing in thermal barrier and the outlet half in heat transfer hehe.

 

[140Airpower]

Actually, the rotors will always be much hotter than the part of the housing they run in. The bore region of the housing will tend not to track rotor temp very well because inlet air is cool and the housing is directly heated only at the bottom.
I have not heard of blowers suffering loss of strength due to temperature even though the compressed air reaches 400F or so. Note that piston tops and air cooled cylinder heads get much hotter and survive heavy service.

 

[yoshimitsuspeed]

The concern is not structural integrity. It's about the rotors expanding enough to contact the housing.

 

[tbuelna]

That Rootes device creates increased manifold pressure by forcing a volume of air into the intake space that is larger than the engine naturally wants to flow. This compresses the manifold air volume constrained between the rotor lobes and the intake valves. The efficiency of a two-lobed Rootes device is normally fairly modest. The main source of efficiency loss with a Rootes blower is leakage past the rotor tips and sides. Sometimes tip seals or sacrificial coatings are applied to the tips to reduce leakage.

With regards to operating temperatures and thermal expansion of the rotors and housing, if both are made of similar materials they should have similar thermal expansion rates. However, with a Rootes blower the discharge side of the housing will be exposed to higher air temperatures than the intake side, while the spinning rotors will have a fairly uniform exposure to the low/high air temps at the intake and discharge ports.

The main problem I can see with higher boost levels in this supercharger is that the higher air temperatures created at the discharge side will result in more thermal distortion in the housing than it was designed for. Most likely, this will increase leakage past the rotor tips and reduce efficiency.

 

[yoshimitsuspeed]

Roots blowers are actually one of the least efficient ways you can compress air.
I'm not at liberty to show the Ogura maps I have but the compressor above is slightly less efficient than a comparable Eaton M62. The Ogura charger is only rated to a pressure ratio of 1.8:1 and has slightly lower efficiencies at 1.8 than the M62 does at 2:1. As you can see, with a roots blower it's quite easy to get into the mid to low 50s in efficiency especially in small displacement motors that require high pressure ratios to make power.

As I already tried to explain several times including the post directly above yours, the issue is not increased leakage or air gaps but quite the opposite. The rotor will tend to heat up more and faster than the housing. This causes the rotor blades to hit the housing.

 

[tbuelna]

A Rootes device can be made to give good efficiency, but it requires close tolerance tip clearances, twisted multi-lobed rotors, etc. I also would not accept that the two lobed rotor in the design noted would experience higher operating temperatures than the discharge end of the housing. While the rotor lobes are more thermally isolated than the housing structure, the rotor lobe surfaces are also constantly exposed to alternating low air temperatures on the intake side and higher temperatures on the discharge side. As for the housing, the discharge side is constantly exposed to higher air temperatures.

 

[Tmoose]

Hi Yoshimitsuspeed,

" The rotor will tend to heat up more and faster than the housing. This causes the rotor blades to hit the housing. "

Have you been able to examine or better yet autopsy any failed superchargers? I'd be interested to see the wear patterns on the rotors and housing, especially some in the early stages of failure. I envision Scuffing of the inner contours of the rotors could be the result of phasing type errors, or rotor-to-rotir clearance set by the center-to-center spacing of the rotors.
Warming up the housing might help maintain rotor-to-case clearance, but not have so much effect on rotor-to-rotor clearance.

I assume greater boost is achieved by speeding up the supercharger.
What is the max rpm of the supercharger in OEM use? How high are the ones that fail revving?

My earlier question about the certain point that failure may ensue should have been more specific.
Are there Boost levels or supercharger rpm that are related to failure probability?
Do the failures happen shortly after high boost is applied, or only after a full minute at full boost?
The higher pressure that exists on "one side" of the rotors would tend to bow the rotors (perhaps insignificantly) and also bias the rotors in their available bearing clearance, although the gear reaction forces may be more significant. 15 psi applied over 32 square inches is about 500 lbs of force.

The folks that build GMC blowers for more severe use talk about being concerned with stuff like securing the rotors to the shafts, and end clearance adjustments that (due to the helical drive gears) change the rotor phasing and thus the rotor-to-rotor operating clearance.

Those same GMC folks talk about using different rotor/case clearances depending on whether fuel is passing thru the rotor mesh. This may support the water injection remedy, in addition to the detonation suppression benefit.
( GMC rotor to case clearance is adjusted by moving the rotor bearing housings up or down in the case, not with manufacturing tolerances )

 

[140Airpower]

tbuelna, ...The main source of efficiency loss with a Rootes blower is leakage past the rotor tips and sides....

The main inefficiency, even in a perfect Roots blower, is that the compression of air is a turbulent process. There is NO compression in the blower itself. The compression occurs with two turbulent flows, -first a burst flow from the pressurized manifold back into the atmospheric pressure volume between the lobe and housing as the lobe opens up to the pressurized manifold. Then, the two lobes mesh to squeeze air from the lobe space and pack it into the manifold, which is done without the benefit of an aerodynamic diffuser. Leakage adds inefficiency on top of this.
In contrast a centrifugal compressor can reach efficiencies in the high 80% range even without compressor to housing sealing because some compression occurs smoothly in the compressor itself while the remaining compression occurs smoothly in a well designed diffuser that converts gas velocity into static pressure.