Another great thing about the DTM!

[speedy57tub]

Jake,
I am still trying to get the oil temps down on my Speedster's 2110-cc. I just got back from driving it around and the oil temperature is still at 260 degrees! I did add a funnel inside the shroud and it did have an affect in that it took noticeably longer for the oil temperature to rise to the 260 degrees. I did notice that the car hesistated between 3rd and fourth gear and I am wondering if jetting is the culprit. I am also a little suspicious that the mechanical thermostat might be allowing some oil to be partially bypassed from the cooler. My #2 CHT is still holding a little shy under 300 degrees. I am going to take out the oil thermostat and then try it again during the week.

After trying to analyze the problem with my 911 setup, I suddenly realized another positive thing about the Joe/Jake DTM design. Because the DTM uses a radial fan, the air is expelled in a tangential manner away from the perimeter of the fan. This is great because the air can be partitioned and directed into the 5 different areas, the four cylinders and the oil cooler. Of course, Joe and Jake have divided up the air not equally, but based on trial-and-error (calculations do not always take into consideration all real-life variables) based analyses. To determine the amount of airflows to allocate one has to take into consideration the velocity changes and air pressure differentials that the individual airflows will encounter on their way to their different allocated targets needing to be cooled. Unlike the axial fan, the higher-pressure, low-CFM radial fans are easier to balance out in one respect. The axial fan designs, like mine are more difficult because the air is expelled in whirling and twisted manner making it difficult to for the air to be readily directed. I've struggled with my design and I am somewhat ok with it but my intent to capture air to cool my oil is not working well. The air coming out of the duct is quite warm and I believe that somehow the heat from the cylinders is being transfered to the airflow. I guess I can use the air to warm my interior and I'll have to maybe use an external oil cooler. I will first try to remove the oil thermostat and see what happens. Wish me luck guys and if any you have any inputs, please forward them to me.

BTW, Jake, please help me out a little, I'd like to move onto the 912! What do you think (besides "I told you so! LOL " ) my next step should be, external oil cooler maybe?

[dstar]

Sell the 356 engine to a collector for $3-$4000, and use it as a down payment on a *nice* type 4 from Jake!


Don

[MASSIVE TYPE IV]

Well there ya have it- finally someone else is understanding why the radial fan works so well! GREAT!

[Guest]

Truth of the matter is that radial fans expel their air tangentially at higher-pressures and lower-CFM compared to axial fans. Axial fans expel with a twist, in our case with a counterclockwise rotation. In reality, both can be used successfully if the proper amount being produced by the fan is allocated properly allocated to each intended target. The problem is that this can only be easily achieved if the distances, changes in directions, air-pressure losses, air velocities, and other variables where all the same for all of the targets. Unfortunately, due to geometrical spatial reasons, the flat-four configuration can more easily be achieved by the radial fan. It is just easier to, by trial-and-error, determine the actual amounts that need to be taken from the radial fan's perimeter to allocate to each target while compensating for the above mentioned variables. The axial fan can also be used but there are a lot more variables that make the design tougher. I was able to make my own 911 design work ok but once I tried to allocate air to cool my oil I ran into some problems. If I was using a radial fan, I most likely would not have a oil cooling problem (assuming that the problem is due to inadequate airflow). Since the axial fan is producing high-CFM but at low-pressures, it seems that the lack of pressure is insufficient to move the air into the funnel within my shroud. Furthermore, I notice that the air being expelled from the flex-hose is significantly warm which means I might be getting some heat transfer from the cylinders and heads! This is most likely due to the plenum enveloping directly right over the whole engine. If the heads weren't so restrictive, the airflows would hae a greater velocity and the heat transfer could be minimized. With the radial fan shrouds, the airflow paths within the shroud are kept somewhat isolated from the targeted areas and only come into contact when the cooling is to occur.

I am still wondering if there is another cause that I am overlooking that is causing my oil temp problem. My CHT for each cylinder are holding at 300 or less. I might have to remove the oil thermostat because I think that it might not be working properly and if that doesn't remedy the problem then I might just have to add an external oil cooler.

[speedy57tub]

Truth of the matter is that radial fans expel their air tangentially at higher-pressures and lower-CFM compared to axial fans. Axial fans expel with a twist, in our case with a counterclockwise rotation. In reality, both can be used successfully if the proper amount being produced by the fan is allocated properly allocated to each intended target. The problem is that this can only be easily achieved if the distances, changes in directions, air-pressure losses, air velocities, and other variables where all the same for all of the targets. Unfortunately, due to geometrical spatial reasons, the flat-four configuration can more easily be achieved by the radial fan. It is just easier to, by trial-and-error, determine the actual amounts that need to be taken from the radial fan's perimeter to allocate to each target while compensating for the above mentioned variables. The axial fan can also be used but there are a lot more variables that make the design tougher. I was able to make my own 911 design work ok but once I tried to allocate air to cool my oil I ran into some problems. If I was using a radial fan, I most likely would not have a oil cooling problem (assuming that the problem is due to inadequate airflow). Since the axial fan is producing high-CFM but at low-pressures, it seems that the lack of pressure is insufficient to move the air into the funnel within my shroud. Furthermore, I notice that the air being expelled from the flex-hose is significantly warm which means I might be getting some heat transfer from the cylinders and heads! This is most likely due to the plenum enveloping directly right over the whole engine. If the heads weren't so restrictive, the airflows would hae a greater velocity and the heat transfer could be minimized. With the radial fan shrouds, the airflow paths within the shroud are kept somewhat isolated from the targeted areas and only come into contact when the cooling is to occur.

I am still wondering if there is another cause that I am overlooking that is causing my oil temp problem. My CHT for each cylinder are holding at 300 or less. I might have to remove the oil thermostat because I think that it might not be working properly and if that doesn't remedy the problem then I might just have to add an external oil cooler.

[Paul Illick]

I am still wondering if there is another cause that I am overlooking that is causing my oil temp problem. My CHT for each cylinder are holding at 300 or less. I might have to remove the oil thermostat because I think that it might not be working properly and if that doesn't remedy the problem then I might just have to add an external oil cooler.

Absolutely. If your CHT's are low and even then you don't really have a an engine cooling problem, but an oil cooling problem. You say you have a 911 shroud? They usually require you to remove the stock oil coolier, so how are you cooling the oil now? What kind of thermostat? Thermostats are made to restrict flow, so I'd bypass it first thing to see if there's an improvement. I'd bet that it.

By the way, you say the pressure from an axial fan is lower than the pressure from a radial fan, but I don't recall ever seeing actual measurements of that. Is that just theory, or was it a direct comparison of the measurements of the pressures in each of those two types of shrouds?

[MASSIVE TYPE IV]

I measured plenum pressures with each shroud I tested. The best 911 system only had 8PSI of pressure while all the radial kits had atleast 10PSI

I measured the pressures with a pitot tube in the upper chambers of each shroud positioned over the centerlines of each bank of cylinders.

[Guest]

Thanks for chiming in guys!
Paul, I went through the painstaking ordeal of designing my own 911 shroud. It was a pain in the arse but, oh well. I now have it on my 2110-cc Type I. To make it work, I had to add some internal vanage to direct the flow. My shroud shape is also somewhat different from those around in that I am not only trying to have the shroud act as a plenum to equalize the internal flow dynamics and pressures (which is rather difficult given the tight spatial constraints and flow impedances) but, I am also relying on the shroud's enveloping shape to induce flow to the #2 and #4 cylinders. I am somewhat satisfied with the results but, unfortunately, failed somewhat in the oil-cooling arena. What I attempted to do was allocate some air via a funnel within the shroud to direct air through a flex hose to a Type IV oil cooler enclosure similar to Jake's DTM oil-cooler housing. But due to the low pressures within the shroud and probably to the lack of adequate airflow velocities through the heads and cylinder cooling fins, there is some degree of heat transfer occurring. I can see this by merely feeling the temperature air coming out of the hose when I disconnect it.

Paul, to answer your question, the differences between the axial and radial fans are “actual” as well as “theoretical”. If one compares the output pressure and flow curves, it is readily apparent that the two different fan designs behave differently and respond in dissimilar manners to RPM changes. If one could have four equal conduit lengths and shapes to the intended four cooling targets, each being supplied by one quadrant of either fan, then we would not be having these axial vs. radial fan issues and debates. We would simply just have to either open or close off the exits to control the flows and pressures for either design and by adjusting the lengths of the conduits, we could normalize the airflows.

BTW, I am a mechanical engineer and I've used some of my past aerospace and building mechanical systems experience to arrive to my current design. The DTM radial fan based design is the simplest and most efficient method for our particular flat-four design. It is just that, given what the airflows have to encounter, it is much easier to harness the air expelled from the radial to direct to the individual cylinders. Due to the dynamic and directional fashion that the airflow is expelled from the axial fan, it is just too difficult within such a small envelope to direct it effectively. Also due to the lower pressures, we cannot influence the air to move like we want it. In the past, Jake and I discussed how easily minor changes within his Type I DTM design influenced pressure differentials within the shroud and how cooling changed from cylinder to cylinder unpredictably (one reason the Type I heads are just more flow restrictive, compared to the Type IV, due to their cooling fins). At the moment, it first appeared as a negative thing but once one understands what is happening, then we realize we can use this effect to rather effectively balance the design. I am sure that is what Joe and Jake have done with the Type IV DTM design.

As far as the 911 shroud is concerned, a functional design can be done but results heavily rely on much many other factors and variables. With all that Jake has experienced and encountered, I would not be surprised if he actually makes his 911 work actually well.

[speedy57tub]

Thanks for chiming in guys!
Paul, I went through the painstaking ordeal of designing my own 911 shroud. It was a pain in the arse but, oh well. I now have it on my 2110-cc Type I. To make it work, I had to add some internal vanage to direct the flow. My shroud shape is also somewhat different from those around in that I am not only trying to have the shroud act as a plenum to equalize the internal flow dynamics and pressures (which is rather difficult given the tight spatial constraints and flow impedances) but, I am also relying on the shroud's enveloping shape to induce flow to the #2 and #4 cylinders. I am somewhat satisfied with the results but, unfortunately, failed somewhat in the oil-cooling arena. What I attempted to do was allocate some air via a funnel within the shroud to direct air through a flex hose to a Type IV oil cooler enclosure similar to Jake's DTM oil-cooler housing. But due to the low pressures within the shroud and probably to the lack of adequate airflow velocities through the heads and cylinder cooling fins, there is some degree of heat transfer occurring. I can see this by merely feeling the temperature air coming out of the hose when I disconnect it.

Paul, to answer your question, the differences between the axial and radial fans are “actual” as well as “theoretical”. If one compares the output pressure and flow curves, it is readily apparent that the two different fan designs behave differently and respond in dissimilar manners to RPM changes. If one could have four equal conduit lengths and shapes to the intended four cooling targets, each being supplied by one quadrant of either fan, then we would not be having these axial vs. radial fan issues and debates. We would simply just have to either open or close off the exits to control the flows and pressures for either design and by adjusting the lengths of the conduits, we could normalize the airflows.

BTW, I am a mechanical engineer and I've used some of my past aerospace and building mechanical systems experience to arrive to my current design. The DTM radial fan based design is the simplest and most efficient method for our particular flat-four design. It is just that, given what the airflows have to encounter, it is much easier to harness the air expelled from the radial to direct to the individual cylinders. Due to the dynamic and directional fashion that the airflow is expelled from the axial fan, it is just too difficult within such a small envelope to direct it effectively. Also due to the lower pressures, we cannot influence the air to move like we want it. In the past, Jake and I discussed how easily minor changes within his Type I DTM design influenced pressure differentials within the shroud and how cooling changed from cylinder to cylinder unpredictably (one reason the Type I heads are just more flow restrictive, compared to the Type IV, due to their cooling fins). At the moment, it first appeared as a negative thing but once one understands what is happening, then we realize we can use this effect to rather effectively balance the design. I am sure that is what Joe and Jake have done with the Type IV DTM design.

As far as the 911 shroud is concerned, a functional design can be done but results heavily rely on much many other factors and variables. With all that Jake has experienced and encountered, I would not be surprised if he actually makes his 911 work actually well.

[dstar]

Got any pics you can post of your shroud?

Don

[speedy57tub]

Yup, I'll post them very soon to this thread or another new one. Unfortunately I have to do it at work since I have trouble uploading from home.

[MASSIVE TYPE IV]

Speedy, I have ONE more revamped Type I DTM that I can sell cheaper for test purposes... Want it?

[A_67vdub]

I wonder if putting a honeycomb air straightener directly behind the fan in a porsche style shroud would help in controling the air by making it flow linearly instead of in a spiral pattern? Any thoughts?

Something similar to this:

http://www.dieselpowerproducts.com/tag


Steve

[speedy57tub]

Sure Jake, email me! I am setup with four CHT senders that I calibrated against each other (they are very close).

[dstar]

I wonder if putting a honeycomb air straightener directly behind the fan in a porsche style shroud would help in controling the air by making it flow linearly instead of in a spiral pattern? Any thoughts?

Steve

We used bleed-air starters for the turbines I worked on
that were a combo axial-centrifical style.

From looking at the design and layout of those, it seems
that all you need to make the Porsche more efficient, is to
bifurcate the ducting directly behind the blades, in dual,
outwardly spiralling tubes and duct those to each bank.

Unfortunately, in my design module, it works best with about 39
inches of space. There isn't enough between the back of the alt
and the body of the bug to do it!

Anyway, guess what you end up with, RADIAL discharge tubes!

BUWAHAHAHAHAHAHAHAHAHAHA!

Don