balidey
|
| posted on 6/5/08 at 11:09 PM |
|
|
Working out supercharger pressure off engine
OK, my plan at some point in the future is to get a turbo charger, rip it apart, use the compressor side of it, mounted to a gear to turn it into a
centrifugal type supercharger. This is basically what Rotrex do with a complicated planetary gearbox, needing its own oil supply and cooler. But its
still geared up using pulleys.
So, I plan on doing a DIY version, but with simpler gearing. My engine will be a low blow, I only want to run at about 0.5 BAR (I think, need to
double check that).
But before I fit it I will need to know what speed my supercharger needs to run at. I can connect it to a high speed motor, and then fit a pressure
gauge on the end of the s/charger outlet, but this will surely just pressurise it and build up pressure against the turbine. So how on earth can I
replicate the engine valves opening, ie, I will need some sort of container with a pressure gauge in it and an exhaust vent, so the container
doesn't fully pressurise. Do you get me?
So, is there some way of working this out? What size container, what size hole? I don't fancy bolting it onto my engine with a huge gear
increase only to find out that its too much and blow my headgasket.
I know in some books I read it says fill the charger with oil, rotate the shaft and measure the output, work out volume, which I should be able to
work out pressure from there, but this still doesn't help me with a centrifugal charger.
So, any ideas? Or can someone point me in the direction of more detailed info / books on how to calculate this?
[Edited on 6/5/08 by balidey]
|
|
|
|
|
Simon
|
| posted on 7/5/08 at 12:04 AM |
|
|
Given that turbos spin at anything from 100 to 200 000 rpm, I'd look for a big truck diesel turbo (you'll only be using compressor side,
so no worries about heat on downstream end). Hopefully bigger turbo spins slower.
You'll need to find specs for turbos but depending on engine, may be able to get away with 10:1 drive!
Good luck with the machining
ATB
Simon
|
|
|
clairetoo
|
| posted on 7/5/08 at 05:42 AM |
|
|
I cant help thinking that if this would work , then some one would be doing it 
I think the turbo bearing would last about five seconds with the side loading that a gear would add , and you would still need to spin at silly revs
to get any pressure at all .
Its cuz I is blond , innit
Claire xx
Will weld for food......
|
|
|
Ivan
|
| posted on 7/5/08 at 06:30 AM |
|
|
It's been done a couple of times but you need a pretty sophisticated geartrain because the rapid speed changes breaks things. For instance when
changing gears the engine speed goes from say 7000 RPM to 1000 and back in less than a second so compressor speeds will go from say 70 000 rpm to 10
000 and back in same time if geared direct.
To calculate compressor speed needed you have to have the compressor curves.
This is all very unnecessary effort with a high risk of failure to my mind as a small, correctly sized turbo will give you a very similar torque
curve at the same or even lower boost pressure with much greater efficiency and hence lower charge temperatures. This will translate into a very
similar driving experience with negligible lag.
Most of the reasons that people think superchargers are better than turbos (no lag) are as a result of incorrect turbo selection.
|
|
|
oadamo
|
| posted on 7/5/08 at 06:35 AM |
|
|
i done this a few years back go onto rs components and type in timing gears. and they sell small belts for them aswell ive forgot which company used
these for theres but i just copyed it. i had to have the shaft ground down to fit the bearing at fellows grinding kingswinford. it cost me 10 quid. i
still have the houseing somewhere ill have to dig it out.
adam
|
|
|
jambojeef
|
| posted on 7/5/08 at 07:10 AM |
|
|
You need a copy of "Forced Induction Performance Tuning" A Graham Bell ISBN 1859606911.
It was £25 in the shops or they did a deal on Amazon for this and another turbo-nutter text book. Might be worth a look?
Geoff
|
|
|
balidey
|
| posted on 7/5/08 at 07:15 AM |
|
|
quote:
Originally posted by Simon
Given that turbos spin at anything from 100 to 200 000 rpm, I'd look for a big truck diesel turbo (you'll only be using compressor side,
so no worries about heat on downstream end). Hopefully bigger turbo spins slower.
You'll need to find specs for turbos but depending on engine, may be able to get away with 10:1 drive!
Good luck with the machining
ATB
Simon
Yep, I have considered larger turbos, the rule of thumb for superchargers is large charger at low speed is more efficient than 'tother way
round.
Machining shouldn't be a problem.
quote:
Originally posted by clairetoo
I cant help thinking that if this would work , then some one would be doing it
.
They are, the company called Rotrex do it (just one example), and charge thousands of pounds, but this is the Locost Builders forum.
quote:
Originally posted by Ivan
It's been done a couple of times but you need a pretty sophisticated geartrain because the rapid speed changes breaks things. For instance when
changing gears the engine speed goes from say 7000 RPM to 1000 and back in less than a second so compressor speeds will go from say 70 000 rpm to 10
000 and back in same time if geared direct.
.
Good point, I did think about putting a slipper clutch into the drive mechanism to help and reduce the mechanical shock. Still at the theory stage at
the minute.
quote:
Originally posted by Ivan
To calculate compressor speed needed you have to have the compressor curves.
Ah, yes. Anyone know where to get Garrett standard info from?
quote:
Originally posted by oadamo
i done this a few years back
adam
Good, glad I'm not the only one to think about making one.
And to add, the car its going in, there are some turbo conversions about, but the packaging of the turbo is quite restricted, and also, no one has
fitted a supercharger yet.
|
|
|
BenB
|
| posted on 7/5/08 at 07:41 AM |
|
|
Wow! There's a project and a half.
Without F1-esque budgets and simulators there's no way you're going to know how much boost you'll get until you try it.
Although there are equations out there will try and produce ball-park figures they're pretty rough estimates....
One of the advantages of a Roots type pump is that it's easier to work out an approx %age boost....
|
|
|
balidey
|
| posted on 7/5/08 at 08:27 AM |
|
|
F1 budgets 
All I need is a ball park, there must be some info out there that says something like 'a garrett t3 spinning at 50,000 rpm will give approx
0.5bar of pressure.'
Thats all I need, just an approximate speed to start with. Then I can do the mixing and matching of pulleys. I just need a starting point, so that I
don't start way too high (or too low) with the gearing and pressure.
As I said, I don't mind making up some test aparatus, a chamber, pressure gauge. But the problem is knowing what volume and what outlet, as it
would need to replicate the valves opening and the cylinders filling
[Edited on 7/5/08 by balidey]
|
|
|
bartonp
|
| posted on 7/5/08 at 09:49 AM |
|
|
Test bed it. Connect up to engine!
Use the high speed motor to spin it up, if you can 'borrow' a variable speed drive/inverter then you can start from zero RPM - ie/
slightly worse than std engine.
Increase RPM gradually, measure boost using std boost gauge. Note RPM.
Job done - or is that too simple?
Phil
|
|
|
MikeRJ
|
| posted on 7/5/08 at 12:38 PM |
|
|
quote:
Originally posted by balidey
F1 budgets
All I need is a ball park, there must be some info out there that says something like 'a garrett t3 spinning at 50,000 rpm will give approx
0.5bar of pressure.'
No there isn't because you are missing a vital bit of information: the flow rate.
Depending on the trim of the compressor the pressure generated for any particular flow rate at any particular compressor speed can vary hugely.
I think you are greatly underestimating the complexity of this job. The gearbox in the Rotrex is complex and expensive simply because of the high
mechanical strain it has to cope with combined with the rather large gear ratio.
Any significant side loading on the turbine shaft will destroy the turbo in minutes, so belts are not really viable, at least not to drive the
compressor directly.
Why not buy an ex-mini cooper Eaton supercharger from Ebay for peanuts and know you'll have something that will work?
|
|
|
Ivan
|
| posted on 7/5/08 at 01:40 PM |
|
|
With respect to spinning it up with an electric moter I would guess that at full boost you will need somewhere between 10 hp to 50hp to do the
job.
With respect to the pressure test you would have to work out the air consumption of your motor in the rev range you are aiming at and size a sharp
edged thin plate hole in the outlet box of your test rig to allow that sort of leak at the pressure you are aiming at.
|
|
|
