I know some people have big budgets for their cars, and mine is pretty small, but for me its about the challenge of doing something and following my ideas, rather than having people telling me I cant do what I want to. So when I decided that I wanted to port a cylinder head for my polo, I decided that I needed to do some research instead of just attacking a cylinder head with a dremel and hoping for the best. So I ended up making a flowbench for £15 (not including the wet and dry vac I already had.
so the idea of a flowbench is to measure the the volume of air that can flow though a port at variable valve openings at a given pressure drop. Now I know it can only do that at a steady flow rather that trying to replicate the dynamics of valve opening and closing etc that happen in the real world.
so I fell across this web page
www.dwyer-inst.com/Products/AirVelocityIntroduction.cfm
and then I began to play. now I based my flowbench on 40mm plumbing fittings as they are cheap and easy to get hold of and larger than the ports I was wanting to test.
so this is my version of two nozzles needed to make a pitot tube (which measures the air velocity and because you know the diameter of the tube, you can then calculate the volume of air movement)
and here it is in a tube (shown with an early very sensitive manometer that I used but then decided that I didnt need the accuracy).
I need two manometer, a small one to measure 12" of water pressure (to measure the air velocity with the pitot tube) and a larger 36" one to measure the pressure drop between the port and atmospheric pressure. I made mine from airline for fishtanks, a peice of wood and electrical cable clips
then I needed to make a valve to adjust the pressure drop, this is like a bypass that lets the vacuum suck in unrestricted air and is opened or closed to adjust the pressure on the larger manometer. I used 25 inches of water, which is just under 1 psi.
and then there is the plate that the pumbing tubing fits to the head. yes its mdf and the centre from a roll of cellotape, but since its bore is the same as a polo engine, its all good (and cheap).
and here is the final set up show testing a head. the two manometers are in the background and plasticine is used to smooth the air flowing into the port. bolted on top of the head is a plywood plate with a threaded adjuster that opens the valve 1mm at a time.
I also made a tiny pitot tube, so I could see the differences in velocity arround the port and give an indication about what I needed to do. its a 1/16" o/d tube inside a 1/8" o/d tube.
of course its not the worlds most accurate peice of equipment, but its fit for purpose. One thing that surpised me with flow testing is the noise, as it was deafening as I needed to wear ear defenders on the standard head. of course noise is vibration which is turbulance, so and the flow gets better the noise goes as the turbulance does.
I had already decided that the ports on the polo head were too large based on articles I had read about engine tuning, and ended up developing the ports similar to the ones below (compared to standard ones at the top of the pictures). as Im too tight to buy casting silcone rubber, I used alginate which is used by dentists to take casts of teeth but is about 1/5 of the price.
the thing that suprised me was how abruptly the flow falls off on the standard port, which I think is due to ports being too large, and the changes in shape impeding the flow and creating low velocity areas which create increased turbulance (tested with the small pitot tube).
the pictures above were early tests,and the final version was similar with a higher port floor (and roof) to increase the short side radius. the port shape was developed by using plasticine in the ports and shaping that. when doing flow calculations, I kept it simple and ignored ambient air pressure and humidity, as I had a standard port I could just measure the flow in that one for comparison. I was only after an indication of what was happening to the airflow, not bothered about having xxxxxx cfm airflow and then being asked to prove it.
I went on to modify the head by reducing the ports with JB Weld, and the head is still in use 6 years later with no issues.
so the first thing to do was to roughen the inside of the ports so the JB Weld would adhere
and then to fill the edges of the ports with JB Weld, which took several days and it was done a bit at a time. show below is a port with JB weld in it and another that has been finished
and from the port throat and as you can see I didnt go for a polished finish
I ended up making my own valve seat cutters as commercial ones were far too expensive, so for the carbide cutting inserts, I used carbide electric planer blades as the were pretty cheap,and they were just glued to some slots in the cutter heads
the valves were also reprofiled,with the modified one being on the right
and the result was that when it was rolling roaded I got 98bhp at the flywheel. ( I was hoping for 100bhp, but that should be possible by tweaking the fueling and ignition).
the previous time the car was rolling roaded it had 88bhp at the flywheel (basically the same setup apart from the headwork). the engine as standard should have 75bph. and I was using stardard fuel injection, standard camshaft, and standard sized pistons and valves, so I guess I didnt do too badly.
when I did this, it wasn't to see how much power I could get out of the engine, it was to improve the driveability as Im a slow driver. what has been impressive is the increase in torque across the rev range, so the engine is smoother to drive and I can trudle about at in 5th gear at 1500rpm (but dont do it too often)
so the idea of a flowbench is to measure the the volume of air that can flow though a port at variable valve openings at a given pressure drop. Now I know it can only do that at a steady flow rather that trying to replicate the dynamics of valve opening and closing etc that happen in the real world.
so I fell across this web page
www.dwyer-inst.com/Products/AirVelocityIntroduction.cfm
and then I began to play. now I based my flowbench on 40mm plumbing fittings as they are cheap and easy to get hold of and larger than the ports I was wanting to test.
so this is my version of two nozzles needed to make a pitot tube (which measures the air velocity and because you know the diameter of the tube, you can then calculate the volume of air movement)
and here it is in a tube (shown with an early very sensitive manometer that I used but then decided that I didnt need the accuracy).
I need two manometer, a small one to measure 12" of water pressure (to measure the air velocity with the pitot tube) and a larger 36" one to measure the pressure drop between the port and atmospheric pressure. I made mine from airline for fishtanks, a peice of wood and electrical cable clips
then I needed to make a valve to adjust the pressure drop, this is like a bypass that lets the vacuum suck in unrestricted air and is opened or closed to adjust the pressure on the larger manometer. I used 25 inches of water, which is just under 1 psi.
and then there is the plate that the pumbing tubing fits to the head. yes its mdf and the centre from a roll of cellotape, but since its bore is the same as a polo engine, its all good (and cheap).
and here is the final set up show testing a head. the two manometers are in the background and plasticine is used to smooth the air flowing into the port. bolted on top of the head is a plywood plate with a threaded adjuster that opens the valve 1mm at a time.
I also made a tiny pitot tube, so I could see the differences in velocity arround the port and give an indication about what I needed to do. its a 1/16" o/d tube inside a 1/8" o/d tube.
of course its not the worlds most accurate peice of equipment, but its fit for purpose. One thing that surpised me with flow testing is the noise, as it was deafening as I needed to wear ear defenders on the standard head. of course noise is vibration which is turbulance, so and the flow gets better the noise goes as the turbulance does.
I had already decided that the ports on the polo head were too large based on articles I had read about engine tuning, and ended up developing the ports similar to the ones below (compared to standard ones at the top of the pictures). as Im too tight to buy casting silcone rubber, I used alginate which is used by dentists to take casts of teeth but is about 1/5 of the price.
the thing that suprised me was how abruptly the flow falls off on the standard port, which I think is due to ports being too large, and the changes in shape impeding the flow and creating low velocity areas which create increased turbulance (tested with the small pitot tube).
the pictures above were early tests,and the final version was similar with a higher port floor (and roof) to increase the short side radius. the port shape was developed by using plasticine in the ports and shaping that. when doing flow calculations, I kept it simple and ignored ambient air pressure and humidity, as I had a standard port I could just measure the flow in that one for comparison. I was only after an indication of what was happening to the airflow, not bothered about having xxxxxx cfm airflow and then being asked to prove it.
I went on to modify the head by reducing the ports with JB Weld, and the head is still in use 6 years later with no issues.
so the first thing to do was to roughen the inside of the ports so the JB Weld would adhere
and then to fill the edges of the ports with JB Weld, which took several days and it was done a bit at a time. show below is a port with JB weld in it and another that has been finished
and from the port throat and as you can see I didnt go for a polished finish
I ended up making my own valve seat cutters as commercial ones were far too expensive, so for the carbide cutting inserts, I used carbide electric planer blades as the were pretty cheap,and they were just glued to some slots in the cutter heads
the valves were also reprofiled,with the modified one being on the right
and the result was that when it was rolling roaded I got 98bhp at the flywheel. ( I was hoping for 100bhp, but that should be possible by tweaking the fueling and ignition).
the previous time the car was rolling roaded it had 88bhp at the flywheel (basically the same setup apart from the headwork). the engine as standard should have 75bph. and I was using stardard fuel injection, standard camshaft, and standard sized pistons and valves, so I guess I didnt do too badly.
when I did this, it wasn't to see how much power I could get out of the engine, it was to improve the driveability as Im a slow driver. what has been impressive is the increase in torque across the rev range, so the engine is smoother to drive and I can trudle about at in 5th gear at 1500rpm (but dont do it too often)
Ingenious adaptations though and a super fast std looking old polo could be a laugh too
