My 3.3L Subie Conversion Thread, Got me some 914rubber stuff Options

[Chris H.]

Yeah there are two ways to do it. 1. "Helper pump" where you keep the other pump and the electric pump just closes the gap when needed and then 2. just having an electric pump only. I'd assume you just take the impeller off with the 2nd way and just let whatever's left spin freely or grind the vanes down.

PRETTY sure you could bleed the system without running the engine with an electric pump Steve. Definitely 90% of the way if you had a simple hose layout. The only thing you don't get is heat, which does help, especially with the last few stubborn air bubbles.

Ray, let me know if you have any questions on the gauges. Given what you've done on your car you are probably more advance than me on the wiring side. It's very easy. You might be able to do what I just did with the coolant temp sender as well. Not sure what the configuration is on the 2.5. The SVX has a sender for the fuel gauge and a sensor for the ECU. Got lucky and the gauge sender was the same thread size as the Speedhut one.

[mgp4591]

Yeah there are two ways to do it. 1. "Helper pump" where you keep the other pump and the electric pump just closes the gap when needed and then 2. just having an electric pump only. I'd assume you just take the impeller off with the 2nd way and just let whatever's left spin freely or grind the vanes down.

PRETTY sure you could bleed the system without running the engine with an electric pump Steve. Definitely 90% of the way if you had a simple hose layout. The only thing you don't get is heat, which does help, especially with the last few stubborn air bubbles.

Chris, would you think that leaving the stock water pump in place would help to keep the water in the engine long enough to complete the heat soak cycle? Just using the helper pump up front to push fluid back from the radiator so the stock pump isn't overworked? My setup isn't running yet so I wouldn't know...
Your build is kickin' it and I'd love to see it sometime, but you know this whole distance thing is a bitch...

[Chris H.]

Yeah any time you're in the area let me know. Steve would be better to answer the electric pump question. I just shared all the knowledge I have (IMG:style_emoticons/default/biggrin.gif) . I'm just running the stock one. It works fine for the street. I think Mike Bellis uses an electric pump maybe? Andrew might? Or is planning to? Someone will chime in.

[mgp4591]

Yeah any time you're in the area let me know. Steve would be better to answer the electric pump question. I just shared all the knowledge I have (IMG:style_emoticons/default/biggrin.gif) . I'm just running the stock one. It works fine for the street. I think Mike Bellis uses an electric pump maybe? Andrew might? Or is planning to? Someone will chime in.

Mike does use an electric pump with the return from the radiator being helped but I think he gutted his stock pump. I hadn't read the post before about your pump being TOO efficient causing cavitation - that's one area where the electric would help out by making sure there's enough water for the pump to draw on.
Also, you mentioned Precision Machine about their project with an EG33 hooked up to an Audi trans? Is that the place in Pennsylvania? I didn't see that listed as one of their projects but that might be a great alternative...

[Chris H.]

No they are in Arizona. Precision Chassis Works. Most of their info is on Facebook

Facebook Photos

Boxster 5 speed trans apparently (Audi):



The headers are a work of art.

Here's a short vid of the car:


https://www.youtube.com/watch?v=Fe_vWc87e1c


Bob has been there. The place is like a candy store for people like us.

[mgp4591]

That looks pretty industrial, damn sexy... I may have to go that way if it's practical and not too pricey! Thanks for the link!

[Chris H.]

Go for the 6 speed if you go Porsche/Audi. Otherwise you might as well do a Subaru trans. Don't need an adapter that way. I thought you were going automatic? The screaming redline shifts are getting to you aren't they? (IMG:style_emoticons/default/happy11.gif)

[motoTrooper]

From what I glean, the guy from Precision Chassis Works cut his teeth doing a 911 eg33 swap in fine fettle.

https://www.youtube.com/watch?v=g4cVMmJKyLg

here's the build thread, really pushed me over the edge on doing an eg33 swap along with Chris H.

http://forums.pelicanparts.com/porsche-911...-3-3-my-sc.html

[Andyrew]

No they are in Arizona. Precision Chassis Works. Most of their info is on Facebook

Facebook Photos

Boxster 5 speed trans apparently (Audi):



The headers are a work of art.

Here's a short vid of the car:

EG 33 swap car

Bob has been there. The place is like a candy store for people like us.

Yup, looks like the 5 speed trans with 100mm flanges. Just about the perfect combo for that engine IMHO.

[914forme]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

SVXs have a wonderful ability to stagnate the pump unless you run a huge hoes into the water pump. We are talking 2" large. This comes into play with high RPM.

Now for the catch, I doubt most people will see the issue in normal street driving. In auto-x where I might hold a high RPMs for an extended period of time, it could be very bad.

Would I make the same decision again, yes I would, I think it is the perfect solution for my car, and combo. YMMV (IMG:style_emoticons/default/confused24.gif)

[rnellums]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

I see this argument a lot on the forums amd I can't for the life of me make sense of it from a heat transfer perspective, maybe someone can help me understand what im missing.

The radiator is in essence a heat sink with forced convective flow through the fins. For a given radiator set up the heat transferred from the radiator to the air depends on only two things, air velocity (and the associated Reynolds and prandtl numbers), and temperature differential between the radiator and the cooling air.

If you are pumping faster, the temperature drop across the radiator will be less, but the volume of coolant cooled will be greater, and since the radiator temperature will be increased on average, the heat transferred out of the system should be greater. This should result in more efficient cooling, and should continue to improve with increased flow with diminishing returns infinitely.

Can anyone poke any holes in this?

[timothy_nd28]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

I see this argument a lot on the forums amd I can't for the life of me make sense of it from a heat transfer perspective, maybe someone can help me understand what im missing.

The radiator is in essence a heat sink with forced convective flow through the fins. For a given radiator set up the heat transferred from the radiator to the air depends on only two things, air velocity (and the associated Reynolds and prandtl numbers), and temperature differential between the radiator and the cooling air.

If you are pumping faster, the temperature drop across the radiator will be less, but the volume of coolant cooled will be greater, and since the radiator temperature will be increased on average, the heat transferred out of the system should be greater. This should result in more efficient cooling, and should continue to improve with increased flow with diminishing returns infinitely.

Can anyone poke any holes in this?

Spoken like a true engineer, even though you are the Purdue type. Go ND (IMG:style_emoticons/default/biggrin.gif)

[A&P Mech]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

I see this argument a lot on the forums amd I can't for the life of me make sense of it from a heat transfer perspective, maybe someone can help me understand what im missing.

The radiator is in essence a heat sink with forced convective flow through the fins. For a given radiator set up the heat transferred from the radiator to the air depends on only two things, air velocity (and the associated Reynolds and prandtl numbers), and temperature differential between the radiator and the cooling air.

If you are pumping faster, the temperature drop across the radiator will be less, but the volume of coolant cooled will be greater, and since the radiator temperature will be increased on average, the heat transferred out of the system should be greater. This should result in more efficient cooling, and should continue to improve with increased flow with diminishing returns infinitely.

Can anyone poke any holes in this?

All I know is if you poke holes in it the coolant runs out on to the floor! (IMG:style_emoticons/default/bootyshake.gif)

[Andyrew]

The electric water pumps (EWP) can either be run by themselves or as a helper. When you do have an EWP you want to remove the thermostat as when the pump is not spinning it essentially wont let any water through if its running as a helper.


If you run a constant return coolant system (Like you should...) You can either bleed the system slowly by turning the key on and letting it cycle at low speed (I dont recommend), you can apply 12v to the pump and get most of the air out of the system, or you can start the car and just fill it up as it warms up (Recommend).

(IMG:http://www.914world.com/bbs2/uploads/post-172-1468264392_thumb.jpg)

My recommendation is that for all 6cyls and below regardless of turbo should run no mechanical water pump unless your build requires you to be below 160deg temp all the time. However most modern engines are efficient up to 225deg this is not a concern.

For V8's especially SBC's I think they might need a high GPH water pump or to run a mechanical and electric pump.

I have been driving my 4cyl around in 90deg weather for about a month now and most of the time its cycling the water pump between slow speed pulses and full and I have it set to not go over 160deg for testing purposes which means that its not even working hard at all to maintain that. I rarely if ever see the fans turn on even putting around. The only time they come on is when the engine shuts off, the water pump and fans turn on until the system has cooled down to 140deg or 2-3mins later. This has an added benefit for me as it cycles the water for the turbo as well cooling it down. Again this is just for testing purposes and to make sure everything is circulating as I flog the turbo.

[Andyrew]

Back to the OP's original video which I watched saturday and forgot the comment. The car sounds great!! Loving that engine combo. I think that should be the new SBC swap as it just make so so much sense in this car.

[jd74914]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

I see this argument a lot on the forums amd I can't for the life of me make sense of it from a heat transfer perspective, maybe someone can help me understand what im missing.

The radiator is in essence a heat sink with forced convective flow through the fins. For a given radiator set up the heat transferred from the radiator to the air depends on only two things, air velocity (and the associated Reynolds and prandtl numbers), and temperature differential between the radiator and the cooling air.

If you are pumping faster, the temperature drop across the radiator will be less, but the volume of coolant cooled will be greater, and since the radiator temperature will be increased on average, the heat transferred out of the system should be greater. This should result in more efficient cooling, and should continue to improve with increased flow with diminishing returns infinitely.

Can anyone poke any holes in this?

Yeah, that forum argument has always bugged me too; it doesn't make any sense at all. Many people seem to think that increased residence time has some ability to increase energy transferred. That's true in stagnant systems since conduction comes into play, but really has no bearing in convection dominated systems.

"Too fast" flow is not a problem for heat exchange since higher than average coolant temperature across the radiator and higher flow rates always correlate to better heat transfer through increased exergy (available energy for transfer) and higher convection coefficients. The air side is really the limiting portion so there are diminishing returns to increasing water flow rate. Ideally there is some optimal flow rate which trades heat transfer effectiveness for pumping losses to reduce parasitic engine loads.

My recommendation is that for all 6cyls and below regardless of turbo should run no mechanical water pump unless your build requires you to be below 160deg temp all the time. However most modern engines are efficient up to 225deg this is not a concern.

Coolant temperatures of 160F are really, really, low. Published studies by Caterpillar have shown engine wear at those temperatures to be very high. IMHO it's much better to be operating at the high end of the coolant temperature limit with a higher pressure radiator cap for safety.

[Andyrew]

My recommendation is that for all 6cyls and below regardless of turbo should run no mechanical water pump unless your build requires you to be below 160deg temp all the time. However most modern engines are efficient up to 225deg this is not a concern.

Coolant temperatures of 160F are really, really, low. Published studies by Caterpillar have shown engine wear at those temperatures to be very high. IMHO it's much better to be operating at the high end of the coolant temperature limit with a higher pressure radiator cap for safety.

This i know. The old issue with the small block chevy's was that if it started to creep up in temp and get over 200 then it just was never going to get back down. Thats why they would put 160deg thermostats on it and run the fans at 180deg.

[76-914]

Electric water pump I am using and Andrew is using is the only pump we have. It actually reduces the flow rate to optimize the engine ability to transfer the heat to the coolant. Yes sometimes things can move to fast or be to big for the combination of systems you have working out.

I see this argument a lot on the forums amd I can't for the life of me make sense of it from a heat transfer perspective, maybe someone can help me understand what im missing.

The radiator is in essence a heat sink with forced convective flow through the fins. For a given radiator set up the heat transferred from the radiator to the air depends on only two things, air velocity (and the associated Reynolds and prandtl numbers), and temperature differential between the radiator and the cooling air.

If you are pumping faster, the temperature drop across the radiator will be less, but the volume of coolant cooled will be greater, and since the radiator temperature will be increased on average, the heat transferred out of the system should be greater. This should result in more efficient cooling, and should continue to improve with increased flow with diminishing returns infinitely.

Can anyone poke any holes in this?

I'm with you. Was wondering the same thing. Could higher volume place some air in the system? Or turbidity that impedes flow? (IMG:style_emoticons/default/confused24.gif)

[mgp4591]

No they are in Arizona. Precision Chassis Works. Most of their info is on Facebook

Facebook Photos

Boxster 5 speed trans apparently (Audi):



The headers are a work of art.

Here's a short vid of the car:

EG 33 swap car

Bob has been there. The place is like a candy store for people like us.

The video link isn't working for me though... The automatic plan is still up in the air but I'll pull some measurements off of cars at RRC here in a few weeks to make my final decision. I'd go with the Audi 6speed but I'm sure they're not giving away the adapters, flywheel, and clutch setups either! If I can get some reassurance that I'm not gonna blow up a Subaru 5 speed I'll probably go that way eventually but if the automatic fits, I'll use it for awhile just to prove it can be done!

[Chris H.]

Fixed the vid! The 5MT Subaru trans is a pretty nice setup for the $. Shifts very smooth. The clutch is light. I don't see how you could blow it up unless you miss gears a lot. I haven't even ground a gear yet!

Andrew thanks for the kudos. Yours is coming along great too. You built that engine right. It's running very cool. I like all the data you get from it.

Hey on the cavitation thing there are lots of threads on the Subaru-SVX.net forum. Apparently the coolant boils because it can't get through the pump fast enough in extreme conditions. Very high revs for prolonged periods. The flow just gets impeded despite the fact that the water pump is working well (too well). It's like traffic waiting around a curve. They even make a high flow water pump inlet with a larger midsection.