Cooling System Mods

[Barry Bordes]

Some thoughts on improving the cooling of the rotor housing in the area of the spark plugs to minimize heat distortion.

Let’s go back to my previous engine failure.
Whenever an engine is disassembled I do a failure analysis (an old habit of mine from working on helicopter turbine engines for a large oil company) and attempt to determine the Root Cause.
Sounds so simple but like they say,” the devil is in the details”.


So first we have the broken apex seal. The very common type failure with the apex seal chip imbedded in the rear rotor and it gouging the housing.

Next we have the cracks at the spark plug holes and lifting of the apex seals in that area.

http://i287.photobucket.com/albums/ll129/bbordes/125-2523_IMG.jpg


Add to that chafing of the side plates to rotor housing interface that same spark plug area.

http://i287.photobucket.com/albums/ll129/bbordes/125-2527_IMG.jpg


At this point we have two contingents drawing different conclusions. One thinks we should secure the area better with dowels and or studs. The other group thinks better cooling of the area is appropriate.

I started looking for 3 studs to reinforce that area. In my search for studs and reamers I contacted Carlos Lopez. He changed my direction.

Carlos thought that creep was a symptom but not the root cause. The problem was high heat and low water flow in that area. He mentioned that the 3rd gen pump was one of the poorest designs.

Kenichi Yamamoto's book shows the temps around the plug area.

http://i287.photobucket.com/albums/ll129/bbordes/Housingtemps.jpg





My thoughts then went to 1-A better water pump (Mazmart) and, 2- Porting the rotor housing in the hot spots as shown above.

Barry

[My5ABaby]

This heat problem also leads to cracks around the spark plugs when someone takes the revs too high on a cold engine. Just a side note... :D

[Phoenix7]

^^very good point BUT Mazda improved on the old rotary designs to minimize this damage in the newer (80's+) rotaries. It IS still possible though.

How would you "port" the housing to improve on the hot spots? I am not able to visualize the modification.

[jgrewe]

Basically you make grooves in the water jacket area around the spark plugs to increase surface area for water to get to.

[Phoenix7]

sounds reasonable, is this common practice (i haven't advanced to internal work so I dunno)?

[Phoenix7]

"Another problem Toyo Kogyo experienced that had been seen at NSU was the formation of cracks in the rotor housing around the sparkplug holes. It was determined that they were a result of "thermal shock" or sudden thermal loads placed on the housing when it was taken from a relatively cool temperature and sharply accelerated to 7,000rpm in all gears on a road test. By developing a test bed procedure that could consistently reproduce these conditions, engineers could then do systematic studies on cooling system improvements and their effect at relieving thermal loads. Tests were showing that on cars driven hard immediately after startup, the temperature of the inner wall of the rotor housing reached 450*F within one minute of startup. This abuse consistently produced cracks in the housing around the sparkplug opening after 4000 cycles.

Improvements in the cooling system brought the maximum temperatures of the rotor housing trochodial wall to below 410*F. But the final improvement was the redesign of the housing itself. The bosses for the tension bolts which held the engine together were relieved from the inner wall of the housing, thereby relieving part of the mechanical restrictions on the part that gets the hottest. Repeated testing after this improvement produced no more cracks in the housing."



So this was back in the 70s. They've made the improvements on the engine since then to further reduce this problem.

You lost an apex seal in the rear housing and I'm under the impression that the rear rotors tend to blow due to heat/preignition (cooling problems you're addressing) but it sounds like you may have also over-boosted and caused the sidehousing damage? :dunno:

Are you monitoring exhaust temps?

Are you going to do the cooling mod and the pump at the same time of will you try the pump first and see if the damage occurs again?


How did you lose the motor (under boost/load)?
What are your mods?
Do you have close up pics of the sparkplug holes?

It's cool that you're documenting and finding things like this. Thanks.

[blwnrtr93]

sounds reasonable, is this common practice (i haven't advanced to internal work so I dunno)?

From what I've seen of water jacket ports they are essentially tiers, to increase the surface area, thereby increasing the heat-sinking capabilities.

[My5ABaby]

So this was back in the 70s. They've made the improvements on the engine since then to further reduce this problem.

Is there any documentation on that? Regardless, it likely still a sizable issue. This is why I get annoyed when people do cold burnouts and stuff. They're just asking for screwed up housings...

[Barry Bordes]

^^very good point BUT Mazda improved on the old rotary designs to minimize this damage in the newer (80's+) rotaries. It IS still possible though.

How would you "port" the housing to improve on the hot spots? I am not able to visualize the modification.

Basically you make grooves in the water jacket area around the spark plugs to increase surface area for water to get to.

Most people do the Racing Beat fin mod around the sparkplug area. This is a big help. It adds surface area and flow turbulence.

http://i287.photobucket.com/albums/ll129/bbordes/RX7_cooling09224.jpg

[Barry Bordes]

The stock 3rd gen water pump with its square leading edges causes cavitation. Notice the damage is at the leading edge area of the housing.

http://i287.photobucket.com/albums/ll129/bbordes/125-2508_IMG-1.jpg


An early attempt at getting the stock pump to not cavitate by contouring the leading edge.

http://i287.photobucket.com/albums/ll129/bbordes/123-2399_IMG.jpg

[Barry Bordes]

"Another problem Toyo Kogyo experienced that had been seen at NSU was the formation of cracks in the rotor housing around the sparkplug holes. It was determined that they were a result of "thermal shock" or sudden thermal loads placed on the housing when it was taken from a relatively cool temperature and sharply accelerated to 7,000rpm in all gears on a road test. By developing a test bed procedure that could consistently reproduce these conditions, engineers could then do systematic studies on cooling system improvements and their effect at relieving thermal loads. Tests were showing that on cars driven hard immediately after startup, the temperature of the inner wall of the rotor housing reached 450*F within one minute of startup. This abuse consistently produced cracks in the housing around the sparkplug opening after 4000 cycles.

Improvements in the cooling system brought the maximum temperatures of the rotor housing trochodial wall to below 410*F. But the final improvement was the redesign of the housing itself. The bosses for the tension bolts which held the engine together were relieved from the inner wall of the housing, thereby relieving part of the mechanical restrictions on the part that gets the hottest. Repeated testing after this improvement produced no more cracks in the housing."

Phoenix7, I would love to read this early research. Even though it is non turbo in its scope.

So this was back in the 70s. They've made the improvements on the engine since then to further reduce this problem.

You lost an apex seal in the rear housing and I'm under the impression that the rear rotors tend to blow due to heat/preignition (cooling problems you're addressing) but it sounds like you may have also over-boosted and caused the sidehousing damage? :dunno:
Actually my rear housing was a RE type which didn't have the latest 3rd gen cooling mod, which contributed to the failure I'm sure.

Are you monitoring exhaust temps?
Yes, front and rear.

Are you going to do the cooling mod and the pump at the same time of will you try the pump first and see if the damage occurs again?
Both

How did you lose the motor (under boost/load)?
Merging into 70mph traffic.

What are your mods?
RE 13B, ITO large street port, TO4S, HKS cast divided, 50/50 AI

Do you have close up pics of the sparkplug holes?
Yes, but they are not the best.

It's cool that you're documenting and finding things like this. Thanks.
I will put pictures soon. Thank you for your input.
Barry

[Barry Bordes]

"Improvements in the cooling system brought the maximum temperatures of the rotor housing trochodial wall to below 410*F. But the final improvement was the redesign of the housing itself. The bosses for the tension bolts which held the engine together were relieved from the inner wall of the housing, thereby relieving part of the mechanical restrictions on the part that gets the hottest. Repeated testing after this improvement produced no more cracks in the housing."

Phoenix7, do you have pictures of how "The bosses for the tension bolts which held the engine together were relieved from the inner wall of the housing". Were they severed completely? The housing wall just floated? For how far?

Very Interesting, Barry

[Barry Bordes]

I actually did two housing mods. The first was to enlarge the passages in the hottest areas next to the sparkplugs as shown in Yamamoto's housing temp diagram.
But use caution! I cut through on the lower side of the Trailing plug. A bad combination of new carbide burs and being too large. 1/4" long shaft would probably be best.

http://i287.photobucket.com/albums/ll129/bbordes/125-2536_IMG-1.jpg

[Phoenix7]

Phoenix7, do you have pictures of how "The bosses for the tension bolts which held the engine together were relieved from the inner wall of the housing". Were they severed completely? The housing wall just floated? For how far?

Very Interesting, Barry
No pics at this moment. I'll look and see if I can find the info when I get home. That excerpt came from the most recent book on the rotary egine (released 2007 or 2008) and this section's source was an SAE paper from 30 years ago. I'll let you know which one (in case you can access the SAE papers) in a bit.

I actually did two housing mods. The first was to enlarge the passages in the hottest areas next to the sparkplugs as shown in Yamamoto's housing temp diagram.
But use caution! I cut through on the lower side of the Trailing plug. A bad combination of new carbide burs and being too large. 1/4" long shaft would probably be best.

http://i287.photobucket.com/albums/ll129/bbordes/125-2536_IMG-1.jpg

I see! thanks, good stuff.

[Phoenix7]

so the info about the cooling is cited from:

"Toyo Kogyo's Research and Development on Major Rotary Engine Problems" SAE transactions. Paper #700079 for SAE meeting of January 1970.

[-xlr8planet-]

Im doing the water jacket mod. Will post pics tomorrow.

Im doing a crappy job BTW. Doing it by hand its quite hard....

it does the job anyways, even if it does not look pretty

[-xlr8planet-]

http://farm3.static.flickr.com/2341/2462477743_0b852a0ede_b.jpg

http://farm4.static.flickr.com/3120/2463313958_2e47f733ac_b.jpg

http://farm3.static.flickr.com/2417/2463326348_767cfda7ba_b.jpg

http://farm3.static.flickr.com/2032/2462489411_da5b795541_b.jpg

[Chadwick]

Another place to look is the flashing left over on the irons from the mold. I go through and clean all of the excess metal off to smooth out the flow.

Dan

[-xlr8planet-]

Just for the record:

to do this properly you gotta have great pulse ... or a CNC machine up your ass

:D

[Barry Bordes]

On my next set of housings I am going to try a wood router and carbide bit. Routers have great height adjustments.

Barry

[Barry Bordes]

This heat problem also leads to cracks around the spark plugs when someone takes the revs too high on a cold engine. Just a side note... :D

I never load the engine cold. I wait for 80˚C before boosting and always use at least a 2 minute cool-down before shutting down (a habit from working on turbine engines which is their requirement).

So I think the cracking has to be just high heat from poor cooling in that area.

The housing with the cracks didn't have the 3rd gen housing cooling mods, but both housings were lifting in the spark plug area. Notice the carbon stains to the side of the hole.

http://i287.photobucket.com/albums/ll129/bbordes/125-2523_IMG.jpg

I have a theory that this lifting of the apex seal and jamming of the triangle tip is what breaks the seal in some cases without detonation being present. I found one unbroken but cracked apex seal. It was cracking from the bottom! I visualized them breaking from the top.

Barry

[Barry Bordes]

I am thinking that we need to increase the temperature in the carbon stained areas and drop the temp at the plug so I am going to try this.

The slotted spoke will concentrate the cooling conduction effect to the plug boss but not to the housing.

Maybe added grooving perpendicular to water flow would also help.

Barry

http://i287.photobucket.com/albums/ll129/bbordes/126-2659_IMG.jpg

[Phoenix7]

damn, i see what you did. So, how safe is it to remove so much of that metal? What boost do you normally see?

[Barry Bordes]

Normally 14 but up to 20 psi. I checked out a 16X and a rx8 housing for the latest cooling mods. They are similar. But they remove most of the fin that could be used to cool the plug bore.

Barry


http://i287.photobucket.com/albums/ll129/bbordes/Rx8Housing.jpg

[Herblenny]

Very Interesting Barry! I would love to see the results of this mod. Maybe we could get one of the CNC guys to program this in and mass cut the housing... I also like the idea of parallel groves vs. the perpendicular grooves I've seen before.

[Barry Bordes]

Very Interesting Barry! I would love to see the results of this mod. Maybe we could get one of the CNC guys to program this in and mass cut the housing... I also like the idea of parallel groves vs. the perpendicular grooves I've seen before.

This will be my next test. Presently I am running the enlarged passages around the sparkplugs with the Mazmart pump.
Barry

[vex]

Another place to look is the flashing left over on the irons from the mold. I go through and clean all of the excess metal off to smooth out the flow.

Dan

You actually release more thermal energy with turbulant flow (not saying removing the flashing is a bad thing, just that it would be better to cause turbulence in the coolant stream so that lower energy coolant comes in contact with more surface area).... did I make any sense at all?

[Barry Bordes]

You actually release more thermal energy with turbulant flow (not saying removing the flashing is a bad thing, just that it would be better to cause turbulence in the coolant stream so that lower energy coolant comes in contact with more surface area).... did I make any sense at all?

Yes it does. I would like to use computer heat-sink type devices on the flat surfaces next to the plugs.
Barry

[vex]

Yes it does. I would like to use computer heat-sink type devices on the flat surfaces next to the plugs.
Barry

Fin type (copper)?

[Barry Bordes]

Fin type (copper)?

To minimize dissimilar metal corrosion we should stick with aluminum.

Barry

[vex]

To minimize dissimilar metal corrosion we should stick with aluminum.

Barry

That's what I was curious about.

[classicauto]

Doing some housing mods for a current rebuild. Just setting up a bit to do the common jacket mod.

I was thinking about the fin idea......whats the thoughts on basically standing some AL sheets in the paths around the plugs? I'm thinking if they're sized properly the sandwiching of the engine oculd hold them in.....? Not entirely sure though, I certainly don't want the twisting around and clogging up a vein.

[Barry Bordes]

My best guess would be to oven braze two aluminum fins into slots cut into the main existing fin. Radiator type solder may work.
Barry

[Cp1]

Very Interesting Barry! I would love to see the results of this mod. Maybe we could get one of the CNC guys to program this in and mass cut the housing... I also like the idea of parallel groves vs. the perpendicular grooves I've seen before.

Haha i would love to do this! i have already written a few programs for milling out rotors and some other things but if anyone wants to donate a set of useable housings i would be willing to program and machine them.

I dont like the parallel grooves. by observation, coolant flow is horizontal and those grooves seem like a great place to trap air. which would be counter productive to cooling. i would much rather see the parallel grooving.

[Barry Bordes]

Haha i would love to do this! i have already written a few programs for milling out rotors and some other things but if anyone wants to donate a set of useable housings i would be willing to program and machine them.

Cp1, how would you propose to fuse the new fins in? Could a piece of aluminum square tubing be used? (harder to machine two grooves but a lot more surface area!)
Barry

[Phoenix7]

can't wait to see the results fellas.

[nissanconvert]

subscribed

[Cp1]

Cp1, how would you propose to fuse the new fins in? Could a piece of aluminum square tubing be used? (harder to machine two grooves but a lot more surface area!)
Barry

Sorry for the delay forgot which section this discussion was in and couldnt find it. but to answer your question i would just mill straight down the width of the housing, so longitudinally front to back if you would were looking at the motor built and in your car. i wouldnt add metal to the housing just take some away in logical places. ive been busy working on this for a little while so i should have something to show soon.

This weekend im going to the wrecker and ill see if i can grab some garbaged ones to use as an example.

[classicauto]

Attached a pic of one of the housings I did for my recent (and first entirely solo) rebuild.

I just did the common jacket mod, nothing super special. The only other thing I did thats neat is welded up the EGR holes in the exhaust sleeves and groudn them smooth. Take a good look at a housing next time you get the chance and take note of all the cavities that would be filled with (at WOT +1600F) HOT exhaust gas and soaking into the coolant as a result.

Don't know how effective that mod will be, but I'm giving it a shot nonetheless.

On the rebuild I'll do on the keg I'm swapping this one for, I plan to do similar jacket grooves (done with a dremel engraver bit on the drill press, nothing too special...) but with my newly found thin and LONG end mill I will add a piece of 1/8" AL to either side of the plug bore on a diagonal. I'll be cutting out a piece from center of the "divider/support" between the dowel bores and then notching the inside and outside edges to provide support for the piece. It will be a minor interferance fit and the sandwiching of the engine will also prevent it from uprooting afterwards.

Keep the discussion/thoughts going :)

[Barry Bordes]

On the rebuild I'll do on the keg I'm swapping this one for, I plan to do similar jacket grooves (done with a dremel engraver bit on the drill press, nothing too special...) but with my newly found thin and LONG end mill I will add a piece of 1/8" AL to either side of the plug bore on a diagonal. I'll be cutting out a piece from center of the "divider/support" between the dowel bores and then notching the inside and outside edges to provide support for the piece. It will be a minor interferance fit and the sandwiching of the engine will also prevent it from uprooting afterwards.

Keep the discussion/thoughts going :)

Sounds good. Can't wait to see a picture.

I wonder if you could fuse the 1/8" AL to the "divider/support" for better heat transfer?

Barry

[howard coleman]

congrats on an excellent thread. i see so many housings w plug hole cracks or carbon evidence of distortion at the plughole...

i really like the longitudinal (Cp1) grooves to add surface area. i also like the idea of welding the sleeveholes closed to help the coolant.

excuse me but i have got to go as i have some welding to do...

this thread is definitely heading in an important direction.

hc

[Barry Bordes]

congrats on an excellent thread. i see so many housings w plug hole cracks or carbon evidence of distortion at the plughole...

hc


Glad to see you over here Howard. We can use your experience and expertise.

Any other ideas on cooling the sparkplug boss area?

Barry

[classicauto]

Sounds good. Can't wait to see a picture.

I wonder if you could fuse the 1/8" AL to the "divider/support" for better heat transfer?

Barry

Didn't think about fusing them barry, but I suppose a tack on each corner wouldn't hurt.....just don't want to warp anything.

I'd imagine though, that a press/interferance fit would still have heat transfer through the contacting areas - but correct me if I'm wrong.


__________________________________________________ ______________
I'm getting ready to set this up here though.........and just wanted to pose the question to you guys on whether or not you think the AL used for the housings is anything super special? (in the areas we're talking about adding to)

Reason being, is it was brought up in a discusssion I had recently that some of the sheet AL I have around the shop, may expand at a higher rate then a cast aluminum part.......if this is the case, it would risky to do what I'm proposing if the piece expands too far and creates more problems then it fixes. If it does pose a problem - what grade of AL should I be looking to use?

Thoughts?

[vex]

Didn't think about fusing them barry, but I suppose a tack on each corner wouldn't hurt.....just don't want to warp anything.

I'd imagine though, that a press/interferance fit would still have heat transfer through the contacting areas - but correct me if I'm wrong.


__________________________________________________ ______________
I'm getting ready to set this up here though.........and just wanted to pose the question to you guys on whether or not you think the AL used for the housings is anything super special? (in the areas we're talking about adding to)

Reason being, is it was brought up in a discusssion I had recently that some of the sheet AL I have around the shop, may expand at a higher rate then a cast aluminum part.......if this is the case, it would risky to do what I'm proposing if the piece expands too far and creates more problems then it fixes. If it does pose a problem - what grade of AL should I be looking to use?

Thoughts?

You'd have to look at the metallurgy book to get the specifics of expansion rates for the cast and cold formed aluminum. I don't think it's anything significantly special as they are cast, and the different aluminum alloys would require more time and facilities than a mass produced cast would.

Also the difference in temperatures isn't going to significantly deform the material to cause interference/failure of the housing.

Using this list from Mechanics of Materials (5th Edition) we have:


Aluminum: Alloy 1100-H14 (99% Al)
Coefficient of Thermal Expansion (10^-6/F):13.1
" " " " (10^-6/C):23.6
Aluminum: Alloy 2014-T6
Coefficient of Thermal Expansion (10^-6/F):12.8
" " " " (10^-6/C):23.0

Aluminum: Alloy 2024-T4
Coefficient of Thermal Expansion (10^-6/F):12.9
" " " " (10^-6/C):23.2

Aluminum: Alloy 5456-H116
Coefficient of Thermal Expansion (10^-6/F):13.3
" " " " (10^-6/C):23.9


Others have similar Coefficients as the ones listed, so we can use it from there.

We'd need to know specifics about the dimensions of the piece being inserted, but looking at the formula of Alpha x DeltaT we would get (say for reference a Delta T= 167F(operating temperature)-75F(room temperature of the measured piece)=92F)
Biggest heat Coefficient of Aluminum Alloy is 5456-H116=13.3 x 10^-6 /F
Plug and chug we get EpsilonT=.001224
Multiple that number by the length of the piece (lets say .5 in) we have .000612 in increase in overall length. Now if we assume that the Housing as the just as high Coefficient and look at the piece directly across the O-ring (about 1/8 in estimate for example) we'd have another increase of .000153 in.
That means you'll have no stress or problems with thermal expansion unless you have the two edges of the piece within .000765 in of each other.

This is just a rough math crunch to work out the numbers and give you an idea of how little the thermal expansion is going to run. Since we don't normally see a DeltaT of magnitudes of order higher than this example I think we'd have the same/similar numbers for the actual piece.

[Barry Bordes]

Here is a product that may work for installing fins on sparkplug ribs.

Controlled Atmosphere Brazing was developed by Alcan, under their trade name NOCOLOK®, as a non-corrosive flux brazing process and is now established as the accepted process for brazing of aluminum heat exchangers.

Barry