Lancer Register Forum banner

Reading compressor maps and matching turbos to your engine.

6.5K views 32 replies 13 participants last post by  Fatman  
#1 · (Edited)
how to Read compressor maps and matching turbos to your engine.

Reading compressor maps and matching turbos to your engine.

A lot of people going for bigger turbos recently so I thought I will take to explain how to read a compressor map for the said turbo. This will show you how to find surge lines, max boost, max power achieved by a turbo. And when you can get (X) boost at ( Y) rpm

First we have to consider the most common engine specs that we have 2 l and 2.3l I will do a rough calucations for both

volume of air (cu ft/min)
= engine rpm x engine cid
------------------------------
(1728 x 2)

This equation is for finding the volume of air going into the engine. The displacement for a 2 l is 122 cu.in and 2.3 l is 140. cu.in . We have a four stroke engine; the intake valve on a cylinder opens once every 2 revolutions of the engine. So, for every 2 revs the engine takes in 122 cu.in for a 2 l and 140cu.in for 2.3l of air. What boost are we running to achieve that will depend depends on the pressure and temperature of the air in the intake manifold. But the volume is always the same every 2 rpm.

So for a 2 l
Vol of air at 3000revs = 3000x 122/(1728x2)= 106 CFM
Vol at 4000revs = 141 cfm
Vol oat 5000revs = 176.5 CFM

For a 2.3 same equation will give you

121 @ 3000revs, 162cfm@ 4000revs, 202cfm @ 5000revs

The Ideal Gas Law what an equation, pretty much rules everything . It relates the air pressure, temperature, volume, and mass (ie, boost in pounds) of air. If you know any three of these, you can calculate the fourth. The equation is written:
PV=nRT
where P is the absolute pressure (not the gauge pressure), V is the volume, n is related to the number of air molecules, which is an indication of the mass (or pounds) of air, R is a constant number, and T is the absolute temperature.

The Ideal Gas Law can be rearranged to calculate any of the variables. For example, if you know the pressure, temperature, and volume of air you can calculate the pounds of air:
n=PV/(RT)
That is useful, since we know the pressure (boost pressure), the volume (which we calculate as shown in the first section "Engine Volumetric Flow"), and we can make a good guess on the temperature. So we can figure out how many pounds of air the engine is moving. And the more pounds of air you move the more power you will make.
Here is the Ideal Gas Law

To get pounds of air:
n(lbs/min)
= P(psia) x V(cu.ft./min) x 29
-----------------------------------
(10.73 x T(deg R))

the reason this is handy is because most compressor maps are actually displayed in lbs/min so this will make easy reference.

Now lets work this out on a car running 1.5bar of boost, Using the figures that we have worked out earlier for volume, for both the 2 and 2.3l configurations. But we will need work out the absolute boost pressure which is the pressure you get on your gauge in psi plus 14.7. so if you are running 1.5bar which is 22psi then add 14.7 = 37.7psi absolute.

Temperature is also need to be absolute which is temperature in the intake manifold in F plus 460. so if you are running a respectable intercooler you shouldn't get more than 45c in the intake that is 113f + 460f = 573F absolute. Now we can apply the equation to get lb/min at various boost levels.

2l first

3000revs@ 1.5bar 18.84lb.min
4000revs@ 1.5bar 25lb.min
5000revs @1.5bar 31.38lbs.min

2.3l
3000revs@ 1.5bar 21.5lb.min
4000revs@ 1.5bar 28.8lb.min
5000revs @1.5bar 35.9lbs.min

I am sure I have lost most people by now but there is light at the end of the tunnel so bear with me.

Now if you look at a compressor map
Image

There are two different sets of curves in the graph; efficiency curves and rpm curves. The area where there are lines drawn is the operating envelope. It is best to operate the compressor within its envelope. It will still run if you go to the right of the envelope, just not well. To the left of the envelope, where it is marked "surge limit", the flow through the compressor is unstable and will go up and down and backwards unpredictably. This is surging. Do not pick a turbo that will operate in this area! It can be very damaging.

Now reading the graph

We will need work correct the flow to inches of mercury instead of psi as most graphs are displayed in that fashion, with out getting into great details here is the equation to use. I know turbonetics use 85F (545R) for their standard temps which is 29degrees so I am going assume that is the standard that all turbo manufactures use. Unless I am adviced otherwise. Pin is your inlet temperation so I am going to assume a 20c which is 70F and also I am assuming a slight vaccum of - 0.5psi in the inlet.

Corrected flow
= actual flow x (Tin/545)^0.5
--------------------------------
(Pin/13.949)

Tin = 70 + 460 = 530 deg R
Pin = -0.5 + 14.7 = 14.2 psia
So corrected flow for

2l figures first
corrected flow for 3000revs
= 18.84 x (530/545)^0.5
-----------------------------------= 18.25lb/min
(14.2/13.949)

4000revs= 24.2
5000revs= 30.42

for 2.3l
corrected flow @ 3000revs= 20.8
corrected flow @ 4000revs= 27.9
corrected flow @ 5000revs= 34.8

So lets take the gt40R that has been used on the MLR recently as an example of an evo turbo with 2.3l
Image


So we mark that point on the bottom of the graph, and draw a straight line upward from that point

Now to get the other point of the graph the pressure ratio Pout/Pin @ 1.5bar or 22pis in the inlet manifold, you we will need to assume a 2psi pressure drop at this level which is standard again so the pressure out of the turbo will be 22psi+2=24 :
Pout/Pin
= (24 + 14.7)
-----------------= 2.72
(-0.5 + 14.7)
So then we find about where 2.72 is on the left side of the graph and draw a line horizontally from that point. Where the two lines meet is where the turbo will operate.
Look at the efficiency curves, which look like circles. Our point is just a little inside the 72% curve, so when we are running at 5000 rpm and 22 psi boost with 20 deg air outside and 45 deg air in the manifold then the compressor efficiency is a fraction over 72%.
All the above figures have assumed 100% volumetric efficiency which is not going to be correct over the whole rev range. This will vary on head design and cams design etc. but if you optimize VE at one part of the rev range you will probably loose out at another area. For the evo engine I would assume the volumetric efficiency to be around 90-95% around 5000revs and probably 100% with cams etc. so the lb/min figures are probably slightly more optimistic. You may need to multiple the figures by 0.9 or 0.95 under 5000revs.

I hope this will be of use for some people, and help understand the general principles of turbo and choice and to a certain respect engine choices as well.

maybe andy should check the calculations just in case i missed something out :) afterall he deals with more complicated equations every day ;) i dont
 
#4 ·
i am glade someone actually liked this thread i was beginning to wonder :D !! but then again my 11 days old son is not having much sleep so i need to keep myself busy during the night.

should be able to make a spread sheet that will work it out from 1000revs to 12000revs.

if you understand this you will be able to work out, what turbo you need, how much to rev it too, what boost is required. etc it is all meant as a guide and not set in stone, if you know what i mean. :)
 
#6 ·
evo-7 said:
i am glade someone actually liked this thread i was beginning to wonder :D !! but then again my 11 days old son is not having much sleep so i need to keep myself busy during the night.

should be able to make a spread sheet that will work it out from 1000revs to 12000revs.

if you understand this you will be able to work out, what turbo you need, how much to rev it too, what boost is required. etc it is all meant as a guide and not set in stone, if you know what i mean. :)
Well done Sam

I look forward to the excel spread sheet been posted
 
#7 ·
Congratulations on the son Sam!

Nice write up, I hope it is appreciated :)

The only deliberate mistake I can see is "We will need work correct the flow to inches of mercury instead of psi" where there appears to be a little confusion of units unless I have missed the point.

I did a previous spreadsheet for a Subaru where the engine size, VE, RPM and PR are input and there is a flow graph drawn which is transparent and overlaid over a compressor plot which is scaled to fit the axes. I have modified it for the GT40 map posted. You can change all the assumptions and the scaling to suit your needs, saves duplicating work. Change the areas in bold. It uses an estimate of power based on 1.42CFM/BHP. This example is for lb/min compressor plots, I may have used different conventions for conversion to standard temperature and pressure, but like all these things it is a guide.

The key thing to getting the spreadsheet going is to size the compressor map to fit the scaling on the Excel plot and get the background transparent, a bit of filling of the grey Garrett lettering and reducing the colours to 16 or even monochrome in paint makes it easier to handle in Excel.

http://freespace.virgin.net/drjohn.banks/flowcalc2333GT40R.xls
 
#8 ·
johnbanks said:
Congratulations on the son Sam!

Nice write up, I hope it is appreciated :)

The only deliberate mistake I can see is "We will need work correct the flow to inches of mercury instead of psi" where there appears to be a little confusion of units unless I have missed the point.

I did a previous spreadsheet for a Subaru where the engine size, VE, RPM and PR are input and there is a flow graph drawn which is transparent and overlaid over a compressor plot which is scaled to fit the axes. I have modified it for the GT40 map posted. You can change all the assumptions and the scaling to suit your needs, saves duplicating work. Change the areas in bold. It uses an estimate of power based on 1.42CFM/BHP. This example is for lb/min compressor plots, I may have used different conventions for conversion to standard temperature and pressure, but like all these things it is a guide.

The key thing to getting the spreadsheet going is to size the compressor map to fit the scaling on the Excel plot and get the background transparent, a bit of filling of the grey Garrett lettering and reducing the colours to 16 or even monochrome in paint makes it easier to handle in Excel.

http://freespace.virgin.net/drjohn.banks/flowcalc2333GT40R.xls
John

Thats a fantastic bit of work

I could use this idea with pump & pipework systems I do

How did you make the cure graph transparent
 
#9 ·
Excellent write up Sam, great minds think alike, thats exactly what i have been working on recently, although i thought my version would be a little complex for the simpletons on this forum. :angel:

I am glad you were able to greatly Simplify my calculations for others to digest ;) :blush:







Edited to say, WTF are you on about :confused: :D
 
#10 ·
Fatman, if you bring up the picture toolbar in Excel, one of the buttons like an arrow selects the transparent colour, then you can have the image in front of the Excel graph and see both. If there are many shades on the background (eg from a scanned image) you need to reduce the number of colours so you have just one background colour you can make transparent.
 
#11 ·
I propose we elect a panel of technical boffins that can us understand our technical problems

We could have a bi-monthley clinic where we could take our sick evos for diagnosis

I would like to propose the following

John Banks
Sam
Andy Fox
 
#13 ·
Excellent thread guys - very informative :D

Can I just ask something really basic :rolleyes: Will running e.g. the GT4088 on a 2.0 produce any less peak horsepower than running it on a 2.3 I4 or a 2.6 I6 (e.g. Skyline GTR) assuming use of all the boost the turbo can make, identical VEfficiency and of same max revs e.g. 8000rpm?

Cheers :)

Rog
 
#14 ·
The Admiral said:
Excellent thread guys - very informative :D

Can I just ask something really basic :rolleyes: Will running e.g. the GT4088 on a 2.0 produce any less peak horsepower than running it on a 2.3 I4 or a 2.6 I6 (e.g. Skyline GTR) assuming use of all the boost the turbo can make, identical VEfficiency and of same max revs e.g. 8000rpm?

Cheers :)
Rog
sounds like your planning something rog? :D
 
#17 ·
the turbo itself will be capable of the same power on any engine.

its weather the engine can follow the required cfm to make this power.

the smaller engine will have to rev higher and run more boost to get the same result as the bigger engine.


i have not done the maths but you probably wont need to rev 2.6l to get the max power out of that turbo.


to make it easier for people to understand, you need to work things backwards. for example you need to decide on your target bhp first, and then choose a turbo that will allow you to run this power, then look at the compressor map to see if you are going to be within this map or not.

a quick example will be a 400bhp evo.

if you want 500bhp then to work out the turob you need

use

airflow in ( lb/min)= hp x afr x bsfc/ 60 ( to convert from hours to minutes)

airflow= 500x11.5afrx 0.55/60= 52.7lb/min

then all you need to do is find a turbo that will make 52lb/min and you know if everything else in the engine is working as it should, you should get 500bhp.

obviously teh 0.55 is an average and the evos can sometimes have a bsfc nearer the 0.52 -0.53 but you get the idea.

lf you still have not lost the will to live by now you can also work out how much boost you will need to run to achieve your 500bhp by using

map= airflowx R( constant 639.6) x(460+temeprature in the intake in F)/ ( ve x rpm/2 x engine capacity in CI )

so for the 500bhp example on a 2l

map = 47.7psi but this is absolute, if you subtract 14.7 atmospheric pressure you will end up with 32psi.

so basically to get 500bhp out of a 2l evo you will need to get a turbo capable of at least 52lb/min and run it at 32 psi which is 2.1bar. which is pretty much the amount of boost i have needed to run in the past on a 2l to hit 500bhp. you may vary by 3-4psi depending on the VE but not by much.

so basically this forumla recommend that the garrett gt3076 is probablky best turbo for the job on a 2l :)
 
#18 ·
The Admiral said:
Excellent thread guys - very informative :D

Can I just ask something really basic :rolleyes: Will running e.g. the GT4088 on a 2.0 produce any less peak horsepower than running it on a 2.3 I4 or a 2.6 I6 (e.g. Skyline GTR) assuming use of all the boost the turbo can make, identical VEfficiency and of same max revs e.g. 8000rpm?

Cheers :)

Rog
at what rpm would you get the peak power on the skyline
 
#25 ·
johnbanks said:
Sam, the 2.0L 3076 @ 2.1 bar example, is this something that 99 RON could support without meth/water/octane booster for running c.500 BHP?
hi john

you can get very close to 500 without race gas, and i mean 10bhp+/- with race gas you can add another 20-30bhp.

the run a very high VE, i would assume the VE of the evo to be around 95%+ at 7-8000revs

neverneverman hit 500bhp on two different rollers with a 2l and gt30r running 2.1-2.2bar