Results of Flow Testing - Improved Intercooler Configuration for Centrifugal Blowers
04-11-2002, 10:40 PM
#1
Results of Flow Testing - Improved Intercooler Configuration for Centrifugal Blowers
Just wanted to let you guys know that the testing went well. The improved configuration flows 37% more air at 28 " H2O than the standard configuration with separate intercooler and ducting (like Procharger). Flow is poorly distributed across the intercooler core in the standard configuration; well balanced with the improved configuration.
This reduction in resistance will allow the compressor to deliver 10% more air to the engine at the same boost, which means 10% more HP. The balanced distribution of air flow across the intercooler should increase cooling efficiency and allow even more boost and airflow, but this can only be proven on the dyno. Dimensions of the prototype allow for up to 650 cubic inches of intercooler core if you really want to go nuts.
I will be talking to the manufacturer tomorrow. Wish me luck!
This reduction in resistance will allow the compressor to deliver 10% more air to the engine at the same boost, which means 10% more HP. The balanced distribution of air flow across the intercooler should increase cooling efficiency and allow even more boost and airflow, but this can only be proven on the dyno. Dimensions of the prototype allow for up to 650 cubic inches of intercooler core if you really want to go nuts.
I will be talking to the manufacturer tomorrow. Wish me luck!
04-12-2002, 08:02 AM
#2
Great and good luck. How is the cooling efficiency? More flow accross the intercooler but how much cooling does it do? Any data on temp in and temp out for the 2 being compared? Any data on psi in and psi out and any cfm numbers in and cfm out? Intercoolers are where the HP and reliability is at.
04-12-2002, 10:31 AM
#4
Marty - These tests were done on flow bench, no compressor, no heat, no water. That's the only way to measure the pressure loss due to system configuration alone. Here are a few of the flow figures for comparison:
Standard Configuration
594 CFM @ 28 " H20
Improved Configuration
816 CFM @ 28 " H2O
Carb Alone (with Flame Arrester)
865 CFM @ 28 " H2O
Carb Alone (without Flame Arrester)
985 CFM @ 28 " H2O
Think about these numbers for a minute. By adding all the pieces you need in the standard configuration to get from the compressor to the carb you take an enormous step backwards, from 985 to 594 CFM. The typical hot rodding approach has been to overcome this resistance by spinning the compressor faster, which requires a larger intercooler to prevent detonation. This is the easiest thing to do and it works. The gains from supercharging and intercooling are so impressive that we forget about the initial step backwards.
But as you know there is always a better way. Without sounding like a smart ass, I have to say that the improved configuration will always be more powerful, because it takes a smaller step backwards to begin with. I too can put on a smaller pulley and a bigger intercooler.
Increasing flow through the whole system is the main thing, but balance across the intercooler core is also important. Cooling efficiency can only be tested on the dyno, but I measured air velocities across the core. This tells you whether the air is trying to drive through one spot in the core, therefore overwhelming the cooling capacity of those fins, while other areas of the core are not utilized and therefore not contributing to cooling at all.
Here are some results for comparison. The results are shown as percentage of total flow in a 4 x 4 grid across the face of the core. Dead zones fluctuate between slightly positive and slightly negative, shown as +/-0.
Standard Configuration
..+/-0...+/-0....+/-0....+/-0..
...4.5....20.1...18.3....+/-0..
...4.5....22.5...30.1....+/-0..
..+/-0...+/-0....+/-0....+/-0..
You can see that the flow is concentrated in the center where the entry and exit ducting are.
Improved Configuration
..4.5....6.3....7.1....7.8..
..6.3....3.2....5.5....8.4..
..4.5....4.5....5.5....8.4..
..7.8....6.3....5.5....8.4..
You can see that even though both of these tests were conducted with the same CFM flowing through the core, the distribution is much better. This means better cooling at any air flow, and it also means that the same size core will do the job at the higher airflow of the improved configuration, since all the core is being used, not just the center.
How's your project coming along. Are we going to be reading about it in Hot Boat this summer?!
Standard Configuration
594 CFM @ 28 " H20
Improved Configuration
816 CFM @ 28 " H2O
Carb Alone (with Flame Arrester)
865 CFM @ 28 " H2O
Carb Alone (without Flame Arrester)
985 CFM @ 28 " H2O
Think about these numbers for a minute. By adding all the pieces you need in the standard configuration to get from the compressor to the carb you take an enormous step backwards, from 985 to 594 CFM. The typical hot rodding approach has been to overcome this resistance by spinning the compressor faster, which requires a larger intercooler to prevent detonation. This is the easiest thing to do and it works. The gains from supercharging and intercooling are so impressive that we forget about the initial step backwards.
But as you know there is always a better way. Without sounding like a smart ass, I have to say that the improved configuration will always be more powerful, because it takes a smaller step backwards to begin with. I too can put on a smaller pulley and a bigger intercooler.
Increasing flow through the whole system is the main thing, but balance across the intercooler core is also important. Cooling efficiency can only be tested on the dyno, but I measured air velocities across the core. This tells you whether the air is trying to drive through one spot in the core, therefore overwhelming the cooling capacity of those fins, while other areas of the core are not utilized and therefore not contributing to cooling at all.
Here are some results for comparison. The results are shown as percentage of total flow in a 4 x 4 grid across the face of the core. Dead zones fluctuate between slightly positive and slightly negative, shown as +/-0.
Standard Configuration
..+/-0...+/-0....+/-0....+/-0..
...4.5....20.1...18.3....+/-0..
...4.5....22.5...30.1....+/-0..
..+/-0...+/-0....+/-0....+/-0..
You can see that the flow is concentrated in the center where the entry and exit ducting are.
Improved Configuration
..4.5....6.3....7.1....7.8..
..6.3....3.2....5.5....8.4..
..4.5....4.5....5.5....8.4..
..7.8....6.3....5.5....8.4..
You can see that even though both of these tests were conducted with the same CFM flowing through the core, the distribution is much better. This means better cooling at any air flow, and it also means that the same size core will do the job at the higher airflow of the improved configuration, since all the core is being used, not just the center.
How's your project coming along. Are we going to be reading about it in Hot Boat this summer?!
04-12-2002, 10:52 AM
#5
All so true. Nice even spread thru the intercooler.
Corky Bell has a new book out on supercharging and briefly hit on this and the need for even distribution thru the intercooler and the use of 'dams' inside to direct the flow.
How about one of those 'Tornado' squirl cage like things at the end of the tube or just inside the intercooler to cause swirl the spread out the air.
It is allways better to run less boost to more power. That is why decreasing all these items will help output- ie- throttle body, air cleaner, manifold, heads and valves, exhaust. Easier in and easier out.
Corky Bell has a new book out on supercharging and briefly hit on this and the need for even distribution thru the intercooler and the use of 'dams' inside to direct the flow.
How about one of those 'Tornado' squirl cage like things at the end of the tube or just inside the intercooler to cause swirl the spread out the air.
It is allways better to run less boost to more power. That is why decreasing all these items will help output- ie- throttle body, air cleaner, manifold, heads and valves, exhaust. Easier in and easier out.
04-12-2002, 11:37 AM
#6
Registered
Join Date: Sep 2001
Location: Houston, TX, USA
Posts: 2,287
Likes: 0
Received 0 Likes
on
0 Posts
TC
From what I am reading with the improved configuration you are gaining on 2 fronts, 1) more area of existing coil being used (before say 200ci intercooler but only 33% is being used so intercollers is 66ci in reality) 2) since air flow will be going thru full area now, air speed is slowed down so more time for transfer of heat. I thing your 10% improvement is way to LOW.
From what I am reading with the improved configuration you are gaining on 2 fronts, 1) more area of existing coil being used (before say 200ci intercooler but only 33% is being used so intercollers is 66ci in reality) 2) since air flow will be going thru full area now, air speed is slowed down so more time for transfer of heat. I thing your 10% improvement is way to LOW.
04-12-2002, 12:26 PM
#7
Turbojack - You may be right, I'm just trying to be conservative until the prototype can be tested on the dyno.
Marty - Turning vanes work but they are a bandaid solution for a poor initial design. The distribution you are seeing is without turning vanes. I can still add turning vanes if I want and see if it helps distribution and/or total flow even more.
Something else to think about. These tests on the flow bench avoid something called system effect factor. What is system effect factor? Simply put the air leaving a compressor discharge is seriously disturbed, and if you ask it to immediately change direction and speed, the actual resistance with the compressor in the picture is much higher than that measured on the flow bench with a nice smooth entry of air from the atmosphere.
The improved configuration is much less sensitive to this system effect factor, so in actual performance testing on the dyno I expect a bigger improvement in air flow than the flow bench results indicate. Coupled with the better cooling efficiency, you might expect more than 10% HP.
Another little thing to think about. If you increase the air flow too much you are going to need a bigger compressor that has its maximum efficiency in the new flow range. That's not a bad thing!
Marty - Turning vanes work but they are a bandaid solution for a poor initial design. The distribution you are seeing is without turning vanes. I can still add turning vanes if I want and see if it helps distribution and/or total flow even more.
Something else to think about. These tests on the flow bench avoid something called system effect factor. What is system effect factor? Simply put the air leaving a compressor discharge is seriously disturbed, and if you ask it to immediately change direction and speed, the actual resistance with the compressor in the picture is much higher than that measured on the flow bench with a nice smooth entry of air from the atmosphere.
The improved configuration is much less sensitive to this system effect factor, so in actual performance testing on the dyno I expect a bigger improvement in air flow than the flow bench results indicate. Coupled with the better cooling efficiency, you might expect more than 10% HP.
Another little thing to think about. If you increase the air flow too much you are going to need a bigger compressor that has its maximum efficiency in the new flow range. That's not a bad thing!
04-12-2002, 01:00 PM
#9
Registered
Join Date: Sep 2001
Location: Houston, TX, USA
Posts: 2,287
Likes: 0
Received 0 Likes
on
0 Posts
Marty
Good point. In looking at my intercooler I coud very easy see that the results would be different if you were sucking verse blowing. To relate to something I know about, on awater pump you put close to the source since you can not suck very far but you can pump long distances.
Good point. In looking at my intercooler I coud very easy see that the results would be different if you were sucking verse blowing. To relate to something I know about, on awater pump you put close to the source since you can not suck very far but you can pump long distances.
04-12-2002, 01:54 PM
#10
Marty - It doesn't matter whether you're sucking or blowing, it's only the difference in pressure from one end of the pathway to the other that matters. That's what drives the air through the pathway. This will not affect the distribution of air flow across the intercooler, it will be bad or good as the case may be whether you are sucking or blowing.
Turbojack - You are right. If you connect the intercooler right after the compressor discharge, the balance across the core could be worse, but look at how bad it is already!
As I tried to describe before, the main difference between the compressor and the flow bench is the mixed up flow entering the pathway from the compressor versus the smooth flow entering the pathway from the atmosphere. This will add resistance and lower the flowrates of both configurations. The improved configuration is much less sensitive to this effect, so the net improvement over the standard configuration will be greater.
It's true that all pumps and fans can push farther than they can pull, but this characteristic has no implication for our discussion of air flow and balance in a duct or pathway. It just means that you want the least resistance possible on the intake of the compressor.
Turbojack - You are right. If you connect the intercooler right after the compressor discharge, the balance across the core could be worse, but look at how bad it is already!
As I tried to describe before, the main difference between the compressor and the flow bench is the mixed up flow entering the pathway from the compressor versus the smooth flow entering the pathway from the atmosphere. This will add resistance and lower the flowrates of both configurations. The improved configuration is much less sensitive to this effect, so the net improvement over the standard configuration will be greater.
It's true that all pumps and fans can push farther than they can pull, but this characteristic has no implication for our discussion of air flow and balance in a duct or pathway. It just means that you want the least resistance possible on the intake of the compressor.
Last edited by tomcat; 04-12-2002 at 06:42 PM.
