The late, great Smokey Yunick came up with the original idea for the benefits of a reverse flow cooling system in the small block Chevy. Eventually, it found it's way to mass production in the LT1 engines.

Here you will find my "test mule" setup for a reverse flow/electric water pump setup for the Mazda Miata, but with a twist.
What makes this setup quite interesting is the fact that the coolant enters the head though it's front and rear sections at the same time.
Meaning, on my original design, the coolant just entered the head at the rear, but that left the front part of the head with "sloppy seconds". So I installed a "Y" at the water pump outlet and now feed both front and rear sections of the head at the same time.

The hot coolant exits the front of the block at the water pump housing. Note that the coolant temp sensor had to be moved to this outlet to properly monitor the temperatures.

During my initial testing, I only went as far as increasing the ignition timing to ~17deg @torque peak. Due to time constraints I was unable to keep advancing the timing to test the limits. However, before the reverse flow setup, I was getting consistent spark knock at settings of ~15deg @torque peak.

There is no doubt that the Miata engine responds well to this system.

Radiator: Honda Civic Aluminum
Water Pump: Davies Craig (I think it is the 80 liter model)

By the way, this car doesn't have a heater core.

 

I am torn here...

On the one hand, this is extremely cool. I've not heard of anyone incorporating a reverse-flow cooling system on the BP, and I can see some huge benefits here.

On the other hand, your intake pipe is wrapped in duct tape.

 

Hey Joe, my whole car is a "test mule"!! Obviously, I am not going to any car shows!!

 

Excellent job innovating on your BP. As you pointed out, it is widely accepted that reverse flow cooling is superior. Do you have before and after CHT readings to compare?

 

i like the idea! but what did u do with the old water pump situation?

 

very interesting setup.I to have read the lit. on Smokeys reverse flow setup and this is very faithful to his design.have you measured the temp in the head?I have a similar setup in mid point of fabrication,and am hoping for comensurate results.Totally cool!(pun).maybe get afabricator to "help",duct tape is a bit much,but then again Ive seen alot of cobbled stuff and this isnt too bad.It looks like it is "free wheeling" the old water pump,correct?Im using a plugged pump with a CNC'd vortex generator
-G-

 

The original water pump is still in the car but is not functioning.

I have already prepared an old water pump (cut and welded) to replace it.

 

I have done temperature measurements on the head, but I have to find where I wrote them. This setup has sat this way for the past 1 1/2 years.

The duct tape is on the intake pipe for IAT measurements...I was testing post intercooler pipe insulation. I was not holding anything together!!

Yes, the original water pump is just along for the ride.

Good luck with your setup!

 

Thanks!I cant get my HUGE pics. to load Or I would show Youz guys what I have going on
-G-

 

Big props for experimentation!!!

I considered doing exactly this years ago, and I came up with a reason why Evans coolant would be recommended for such a setup. I couldn't remember what it was and I had to dig for my own posts:

MX-5 Miata Forum - Skunk2 trick: Electric water pump
MX-5 Miata Forum - electric water pumps ?
MX-5 Miata Forum - Electric Water Pump Replacement

1) A standard mechanical water pump can deliver enough pressure head, (working against the thermostat resistance), to pressurize the coolant, which adds to the cap pressure, to raise water's boiling point which helps prevent hotspots in the head. The Davies Craig electric water pump doesn't, and thus the system would benefit from Evans coolant's high boiling point

2) Evans won a patent lawsuit against GM's use of reverse flow. The patent pertains to redesigning the head coolant passages to prevent steam pockets from forming. This would be a non-issue with the use of Evans coolant. I don't know how much of an issue this would be in a miata head.

The reason I didn't go ahead and do this experiment were:

1) A 2nd pump would be needed for the coolant circuit.
2) My cooling problems were solved by the TDR i/c



To build on sixshooter's points - you could install one CHT probe, then reverse your coolant flow by swapping the in and out hoses of your Davies pump to get before and after readings.

 

what piece did u use here to replace the mixing manifold?

 

Did you steal this idea from Hyper. Is this the "fully disclosed" coolant reroute idea????

 

He has no heater, so all one would need to do is block the inlet from the heater. What he did was use the inlet to mount the coolant temp sensor.

BTW mrtony did you use proportional control for driving the pump? (speed up pump at higher temps)

 

i want to replace the entire mixing manifold, i want to have it pointing forward.

(not sure if its different for a 1.6 or 1.8 but i have a 1.6)

 

 

Thanks!

There is no comparison in flow/pressure from the mechanical pump to the electric pump. The electric pump has the mechanical pump beat in every respect. When you see the flow abilities of the Davies Craig pump you become a believer. You can't run a proper reverse flow system on the Miata with the factory mechanical pump...not even spinning it backwards and with a reverse rotation impeller.

By the way, there is no need for a second pump.

 

No, the pump is driven with constant voltage.

 

How do you regulate temperature?

Check out the links I posted. I *did* investigate the DC pump characteristics. It doesn't have the same pressure/flow capabilities as a mechanical water pump - IOW for a given backpressure (from the t-stat), it can't flow the same. It can't, unless you use a much higher power motor than the 8A they provide. The solution in the DC setup is to run without a t-stat and control temperature by modulating the voltage to the motor. Which leads to the lack of pressure head problem solved by using Evans...

Without a t-stat, and in a system where the main pump can be off or slow during warmup, one would need a separate, smaller pump for the heater circuit.

 

I can set the temperature that the cooling fan comes in with the Link ECU. Max temperature is ultimately control by cooling system capacity/efficiency.

Again, there is no comparison in the pressure of this electric pump to the original mechanical pump. The electric pump is superior in every respect.

 

If the car is moving, the pump is running, there it no t-stat, and the fans are off, the engine can still overcool.

The electric pump is a lower power pump than the typical mechanical water pump. I'm not saying it's "inferior" - I'm saying it needs to run without a t-stat, and a consequence of this is the lower pressure in the cylinder head - which may mean Evans coolant would be needed to counteract this downside. I measured the actual coolant pressures years ago, and I found that when an engine is revved, the pressure rises significantly, even with a full open t-tat. If you look at the pressure-flow curves of the DC (I had to ask for the curves years ago), it can't build more than a couple of psi.

Second, I ran calculations of the required coolant flow given a certain amount of heat rejection, the DC pump cannot work with more than a certain amount of pressure head.