Supercharger Project

okay this is a yes/no answer. from the outset it would seem that the HP is going to be based on the CFM, and you'd be right execpt that the thermal effeciancy of the unit producing the CFM plays a large part in how much you need and where you get it. all engines have pumping losses and so do superchargers. the calculations i do take into account the pumping losses of each with it's thermal effeciancy and 'spit' out an answer. i'll go over the calcs in a moment

i choose 8krpm as the limit for the boost so as to avoid detonation, i can bring the limit down but you'll run into problems with a centifugal unit. a twin screw can build the pressure faster and you can bleed it off at a certain point with fewer pumping loses. the centrifugal would be building the boost past say 6k but now you're bleeding off the boost to keep from going to high and your spending more HP to keep it lower than if you just let it build.

the calcs for the two vary in a few points and this will cause the difference. the twin screw charger is not as thermally effcient as the centrifugal. the higher internal compresion ratio of the twin screw causes more heat and power to drive at low rpm, centrifugals do not have this problem as they have lower internal compresion due to a lack of fixed displacement. this difference is where the twin screw will have loses at both low flow and max pressure. the volumetric efficiancy is greater at low pressure but drops as it rises towards 30psi 95%~80% range the centrifugal is 100% accross the range but due to drive losses it drops to 90% and stays there becouse the there is no fixed volume. the volume increses/degreses with the square root of the compressor rpm. this will also mean that the centrifugal will develop power similar to the NA engine as it increases with rpm in a close linear fashion.

the twin screw is not linear it will produce its max by at least 3krpm this is what makes them great for low rpm gains, the trade off is more heat and the need to shunt boost at cruise and high rpm. the greater force needed to drive the unit at low rpm also take power from the engine. but as you are making so much more it really isn't noticed. except for in your calcs where you will choose the unit you need.

the calcs, i have a few spreed sheets and at first glance it's going to look real confusing. i put these together from the calculations so i wouldn't have to keep grabbing my calculator every time i wanted to see a new change.
the first calc sheet is for the centrifugal and the second is the twinscrew. first we need to know what our pressure ratio is going to be based on our target hp. in this case it's 1.59, from that we can calculate our temp gain. here is the first difference, the centrifugal has a 75% thermal efficiancy vs 70% for the twinscrew. this will start the change, we can see that the twinscrew is heating the air 40*F hotter for the same Psi. this will change the density ratio, which means the twinscrew will lose 21% of it's boost to heat that is horrible an intercooler will be needed to get back those loses. next is the pressure ratio, this is the crux of it. with a thermal difference and also a volumteric one the pressure ratio will be different between them. the PR will be used to calculate CFM needed. now we can calculate for every revolution on a NA engine for CFM. the first calc is just that, the next is the CFM needed based on the pressure ratio we just calculated. becouse the PR is higher on the twinscrew we will need to pump more air for the same HP gain due to heat loses on the twinscrew. what you don't see but would if you kept at the calcs is that the twinscrew will produce this much much sooner. hence the low boost advantage. the centrifugal will drop it's pressure ratio as the rpms drop, we can calculate the max setting but all point in between are based on the unit you choose. the trim and volute will affect the rpm range.

i really hope that make sense, if you want i can send you a copy of the spreed sheet and you can fiddle with it.
Image Unavailable, Please Login
Image Unavailable, Please Login

 

I think I'm just not getting my point across very well. Here are the compressor maps for the MP90 roots blower I used with my first set-up. I ran a drive ratio of 2:1 so you can pull the number off the maps in a form that makes sense to you. It made about 9.5 psi at 6800 and 10.25 at 7700 with the inlet air temp 260+ F...260 is as high as my sensor would read. Remeber my engine is a QV.

The second key point is that the engine flows more air at 6800 than it does at 7700. Even if your number are right and it will flow 40 lb/min at 7700 and 10 psi, it's flowing 44-48 at 6800, dropping the boost to 6psi. That means the HP calc is wrong. The right yeild is about 288, well under the goal. It's works out better with the V1 compress becaue the flow lines are flatter than the V9 in that range.

I don't have the maps for the current crop of twin scew units, but they claim 98% VE and 80% AE, so better than the best centrifugals, not that it matters.
Image Unavailable, Please Login
Image Unavailable, Please Login

 

okay if your talking roots 'roots' than that Ev is terrible. hot air is far less dense than cold air at the same CFM. that is the big difference, the more thermally efficiant the unit the more hp per CFM. twinscrews are better than roots and still less efficiant than centrifugal, however the twinscrew produces boost sooner, that is the advantage. remember there is no way get a twinscrew to exceed the efficiancy of a centrifugal becouse the twinscrew is a fixed displacement charger. and it's difficult to compare them to each other due to the diffeerences in which they generate boost.

if i recalc the peak boost for 10psi at 6800rpm the plot moves left, further from the efficiancy islands. that would drop the V-1 into 68~70% but the V-9 would still be at 74%. as the rpms drop so do the CFM. at the projected 325HP at 6800rpm the Lb/min drop to 33.45 still a PR of 1.6. as we move towards redline the PR moves to 1.90 or 13psi

calculations for each of the units is different.

 

Scott,
I still don't think you're getting it. The positive displacement units are rated in INLET cfm. That's CFM at ambient conditions, efficieny is not in the equaltion yet, it's the same for any compressor. To get 9.5psi manifold pressure at 6800 rpm, I needed about 620 inlet cfm, which is about 46 lb/min of air. Now because the efficiency is low on a roots, the temperature was about 260F. With a more efficient compressor that had a discharge of say 180F, the manifold pressure would drop to about 8.4 psi, but it would still be 46 lb/min and 620 inlet cfm of air.

I you want, Monday I can post the compessor map for the compressor that made the dyno graph I posted, so you can compare flow in any form you choose to measure it to boost.

 

mark, what you are saying for lysholm type compressors is true, however it doesn't work the same for a centrifugal. the pressure ratio on which you multiply the CFM by is temperture dependent. it has no ties to CFM, the CFM is calculated by the PR which is dervied from the power and efficiancy of the charger. i can have two different PR for the same HP due to efficiancy, this will change the needed and generated CFM. that is why a centrifugal, roots and twinscrew all vary on the lb/min needed for the same HP. a centrifugal compresses the air in the volute, a twinscrew compresses the air partly in the chamber and partly in the plenum, a roots simply 'stacks' the air in the plenum - no compression.

a centrifugal compressor is'nt rated by the same inlet CFM becouse each wheel size & trim along with the volute dia, will vary the output based on rpm. that is why the maps are so important for these types. same goes for turbos.

i'll try and explain better, a twinscrew compressor has a fixed volume that is compressed per each revolution. as the rpms increase the CFM does also on a fixed rate becouse the volume in the chamber remains constant. a centrifugal compressor has no volume, it compresses air as the rpms increase and the volute dia. which creates a venturi that allows expansion. this expansion creates pressure, the pressure increases by the square of the shaft rpm.

it's for this reason a twinscrew generates heat at idle and a centrifugal does not. air resists being compressed and generates heat when compressed, the fixed displacement nature of the twinscrew will generate more heat than a turbo/centrifugal compressor. this heat affects the pressure ratio, less dense air at the same CFM equals a loss of boost and a higher PR. a twinscrew has the advantage of the lower rpm range and the centrifugal the higher rpm range. as rpms increase the twinscrew drops in effeciancy, thus the greater psi needed to get HP you want. a centrifugal has better efficiancy as rpms increase, boost builds on the sqaure of the rpms becouse the 'displacement' is increasing.

 

Scott,
The outlet cfm is a very very hard number to work with any compressor. Let me try again.

HP is produced as a direct result of air/fuel mixture being burned. A bigger fire in the same engine must make more hp. The size of the fire is directly related to the mass flow of air and the mass flow of fuel is (or should be) directly related to the mass flow of air.

It does make some difference where the air comes from due to parasitic losses. A turbo eats something like 5%, a centrifugal or screw SC something like 10%, a roots maybe 12%, something like that with all delivering equal mass flows of air. Now here's the part I think you're getting all wrapped around, the boost pressures or PR will not be equal unless the compressors are identical efficiencies. The higher the efficiency of the compressor, the lower the temperatures rise and therefore the lower the boost pressure for a given mass flow.

What type of compressor is used to generate the mass flow doesn't come into the calculation anywhere when you are talking about a single data point, say 6800 rpm. The compressor type comes into the equation when you pick a second point, where that point is relative to the first point, and what the engine is doing at the second point relative to the first.

Efficiecy is efficieny, again it doesn't care what type compressor it is. A screw type will generate heat at low rpm, but because it is compressing, a centrifugal does not, because it is not compressing. At high rpm, the compressor with the higher AE rating will make the least amount of heat and require the lowest shaft power for any given mass flow of air. In order from low to high they go root at about 40-50%, centrifugal at about 70-75%, screw 70-80%.

There are 2 reasons my flow data don't match what you calculate. The first, which I just thought of is that I switched to EFI making my base hp # is about 260hp, so my boost is a bit lower at the specified mass flow rates then a stock engine would be. The second is blow-down.

Blow-down isn't in the calculations anywhere because it is assumed negligible, but on a 308 it is not negligible. About 30% of the air that goes into the engine ends up going straight out the exhaust pipe unburned. That happens because ferrari uses a very low lobe separation angle, long ramps on the cams for the lift they run and the exhaust flow rate is very high compared to the inlet flow rate. Together they add up to an engine that pumps very well, but still has a relatively low VE…because a lot of air gets pumped into the exhaust. The compressor needs to be big enough to deal with the blow-down or it just won’t make boost. I'm going to leave it at that, let you do what you think is best on your system, and not speak of this again.

Getting away from theory, it occurred to me that since you are going blow-thru there is going to be at least 1 psi, probably 2psi going from the blower to the manifold. That means that if the engine will tolerate 10 psi, in the manifold, you’ll need 11 or 12 at the blower to get it. The numbers I supplied are all manifold pressures.

On an engine that makes peak hp well below redline like a 308, and fitted with a centrifugal compressor, a pressure relief valve is about the only way to make good boost at the hp peak without obscene boost at redline. The other solution is a variable ration drive, and while elegant, it is a bit complicated and expensive.

 

let us set aside the 'blow by' waste, i'm not dissagreeing with that. and i don't whish to argue with you, my appologies if it comes across that way. i enjoy the technical debate and i'm sure others are watching in interest. i approached this project as an engineer and went looking for the calcs needed to choose the right unit.

the amount of HP an engine can make is based on density of air to fuel. you can have two airflows of the same CFM but far different densities, the one with the lower density i.e. hot air will produce less power than the higher dense cold air. calculating the CFM that exits the charger is done with testing hence the maps for the centrifugal units. a roots or lysholm is calculated based on displacement and rpm. the thermal effeciancy of each is different and that is what is calculated into the equation to get a pressure ratio, the pressure ratio will be different for each of the 3 types for the same HP goal. this is why they have different CFM requirements. in order to overcome the thermal loss you need to push more air or lower the temp with an intercooler.

when we calculate the CFM of a NA engine it's a simple calc becouse we are dealing with 1 atmosphere. it can be further refined if your over 7k ft. CIDxRPMx.5xEv/1728 = CFM. now when we want to increase the pressure via a charger we need a pressure ratio. the simple pressure ratio is existing HP/desired HP. however this is not accurate as it leaves out the thermal losses for each type of charger. in order to get an accurate Pr we need a density ratio, so we calculate the tempeture gain based on the thermal efficiancy of the charger. at this point we can see the difference in chargers by the density difference. to further seperate them the lysholm chargers have fixed displacements so they operate at a Ve of 92% where as the centrifugal is 100% due to it's lack of displacement. the creates a VeR of 1.25 for the centrifugal and a 1.15 for the lysholm.

in the end this will create three different Pr's and as such when plugged into the CFM calculation the CFM requirement will vary per charger selected. i did another set of calcs. one for each of the units and these are without intercooler or blowby loss.

and i would be very interested in accutual data from your dyno runs as it would allow me to calculate the accutual Ev and losses.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Scott,
It looks like you are using a constant of 14 ft^3/lb and that is where the problem is coming in. If you change temp or pressure, then the ft^3/lb change. There are charts where it's all worked out, but just adding a P1/P2*T1/T2 to it should get you really close.

Burning a fixed amount of air/fuel MUST produce a fixed amount of power and tghe comverse is also true, to make a given hp, a fixed mass of air/fuel is required. A calculation that yeilds a solution other than that must contain an error. The energy must balance, Ein=Eout.

The second error is the engines NA air requirement, there is a term missing. It needs either a pumping term of a loss term to account for air that is pumped, but is not burned because it escapes before the exhaust valve closes. On most engines that number is very low and can be ignored, but on a 308, it's big and goes up as boost goes up.

 

the constant for converting CFM to Lb.min in my calcs doesn't need to account for thermal losses as those were calculated in the density ratio. if you notice the PR increased due to the thermal loss. under a no thermal loss equation i would need less lb/min however the density ratio has changed due to thermal losses. the density ratio affects the lb/min. i sit at sea level so i'm not accounting for atmospheric pressure loss, but that can be accounted for in 1kft stages.

engines burn lb or mass of air as you have noted, the laws of thermodynamics prevent a direct E/in~E/out. there are losses to heat, friction. if you add up the entropy of those conversions then in a gross term they equal. but to get the resultant hp from a specific lb of air we account for temperture and pressure using boyle's ideal gas law and Ve of the engine. all these equations are in my calculations, please look again.

as to the loss of air by blowby, it would have the effect of an increased Ve but the inverse loss of a lower Ve. it can be calculated as long as i know what the loss is.

 

It looks like you have the pressure accounted for, but not the temperature because i don't see it and as soon as I plug it is, all the mass flows are within 1 lb/min of each other like they should be.

Ein=Eout always, that is the 1st law of thermodynamics. The out isn't always what you want it to be (friction and such), but it's alway there. What's true is that the process is not reversible because of friction and such, that's the 2nd law if I recall. With equal mass flows, all the superchargers options will produce equal grossoutput hp from the engine, but the net hp will vary by the efficiency of the compressor with the lower efficiciecny unit yielding a bit less net hp due top higher shaft work required to turn them.

Blow-down is not included in VE. VE is the air/fuel mass sealed in the cylinder for combustion divided by the mass/fuel mass that should fit in the space at the give ambient conditions. Blow-down shows up in mechanical efficieny in that pumping air requires shaft work, but air that escapes unburned does no work. It would show up in a full Ein=Eout analysis (with is really really hard to do), but it is not in VE.

 

the temperture is accounted for, the calcs are right there at the begining. the first step is to calculate the ideal Pr, that is then input into the temp calc from there it is used to calculate the density ratio. that ratio is used to calculate the new Pr based on thermal losses. once we have that new Pr it is fed back in at the top to accuratly calculate the temp. from there it feeds down again and we have the correct Pr and lb/min.

Ve is the ability or lack thereof of the engine to pump air. blowby would affect the Ve as volume is lost during compression. how to accuratly calculate it would be to figure what volume of air we are loosing and adjust the engine displacement to account for it. however i have a feeling that it isn't a fixed loss but linear so as rpms build the loss ratio drops. thus making it really hard to calculate.

take two compressor maps and plot the same points on each, the same Pr & lb/min on one will be different on the other. we have the same air flow but the two will have far different efficencies. one could be right at 75% and the other off the surge end. so gross or net they will not be the same. you can't have the two produce the same gross power, how could it? the less efficiant one would produce less hp. and even this is not an accurate way to describe chargers as they really only increase the Ve of an engine thus multiplying the torque. hp is just the function of it as we know.

 

You're accounting for the temperature rise altering pressure to calc to allow you to calculate pressure correctly. So, when you calculate mass flow, you are using the correct pressure. But the mass flow calc requires 3 variable, CFM, Pres, Temp and you don't have the temp plugged in. At the end, you need to correct the cfm for the temp rise (you already have it corrected for press) or vary the cfm to lb/min conversion factor accordingly. Try it, you'll see the mass flows are the same for each compressor.


VE, like any efficiecny is what you got divided by what you expected. It is the amount of air you are getting into the cylinder divided by what should fit at ambient conditions. Air that goes out the exhasut pipe in by definition not in the cylinder and therefore not part of VE. Air that doen't get into the cylinder due to an intake restriction is basically acconted for in VE, because it need gets into the cylinder, in this case the VE and Pumping Efficiecny are reduced. However air that goes in and staight though, does lower the VE, but was pumped so the PE stays high, this is the case with a 308. It pumps more air than it's low VE would lead you to believe because it's PE is high.

To calculate the blow-down, you simply use the data point I gave you. The measured flow (I gave you) minus the theorectical flow (what you calculated) = the blow-down.


When you are looking for a mass flow, the only thing the compressor maps do at a given point is tell you what rpm to spin the compressor and what temperature to expect the discharge to be. It makes no difference where the mass flow number comes out on the map as long as it is in fact on the map showing the compressor is cabable or delivering it

The maps become important when you try to estamblish a curve for mass flow at every rpm based on the initail point you fixed on the map. The other place they are critical is to allow you to determine what mass flow you will get at conditions the engine will tolerate. So far we have been talking about hp at a given mass flow, but the truth is the engine very much cares about pressure and temperature to keep it from detonating. The maps let you pick a max mass flow to stay under the detonation theshold of the engine.

 

Mark, let me take a moment to address the readers.
mark and I are discussing the temperature that alters the density of air. a engine has a specific volume and based on valve timing and flow charactristics it will alter the volume effeciancy. under atmospheric conditions the air will try and fill that volume under the pressure of atmoshpere or lower depending on your elevation. an engines volume never changes, even under boost. the energy output of and engine is based on air density and fuel density. the commen acceptance is air volume, more air in = more power. partly true. the more air is in reality more air molecules not volume as we can not change that.

air by it's nature resists compression, compressing air creates heat. this heat will lower the density of the air. by adding a charger we are in effect increasing the density of the air in the fixed volume of the cylinder chamber. in order to correctly calculate how much air we need to generate the HP we want we need to take into account a few things. pressure, temp and flow, i have attached another spread sheet but with the calculations spread out so we can all see what is going on. sorry for the small text. size limits and all.

Mark,
i think we may just end up dissagreeing on this point. i have checked with other engineers and i have the temp properly accounted for in the calculations. if i had not calculated the density then yes i would need to calculate the Lb/min based on temp. but the Pr would be off as it would be to low.
Image Unavailable, Please Login

 

Scott,
Dissagreeing is for politics. You have a 1st princilple equation that isn't balanced, that makes it wrong IMO.

Everything you do is correct up until the last line where you calculate lb/min. You are dividing your CFM (which is correct as far as I can tell) by 14 to get lb/min. The error is that 14 applies at about 80F and you have calculated the Pr and therefore the CFM based on hot air, and you even specify that in your step 7, that means 14 doesn't apply until your correct the CFM back down to ambient temp.

I guess in the end it doesn't really matter, the correct lb/min is 32 vs your 34, which really needs to be more like 45 due to blow-down...so your error is in the right direction anyway. The only thing you can't do with the error in you calculation is use the data I provided, which really only means an extra iteration on pulleys size since the compressor you're looking at are big enough to flow the air you're going to need, once you get the right dirve ratio.

 

Here's another way to try the math...and it's a little easier than the way you're doing it.

You want 325hp, you haves 205, so you want 1.585 times more power.

To get it, you need 1.585 time more air. You calculated 280 scfm, so you need 443 scfm to make 325hp.

443/14= 31.7 lb/min, just like you get after all your calculations and the temp conversion you're missing.....but it's still wrong because of the blow down.

 

Or better yet, start right of with the mass flow.

280 cfm = 20 lb/min

20x1.58= 31.7 lb/min of air to make 325 hp

That's as simpe as it can get .

 

A Mondial owner asked me to pop onto this thread and see if I could help.

I have designed, developed and installed supercharger kits on every Porsche 928 from 1978 to 1995, many of which are CIS, K-Jet cars like your 308. Some have said I was the first to ever supercharge the CIS fuel system (I do not knwo about that) but I do know that we are the only ones currently supercharging CIS engines.

Do you need some help with this? I should be able to keep you from making some simple early mistakes in your design.

I'm located in Horicon, WI (about 1 hour North of Madison, WI) and if somebody would like to drop off there 308 for the winter, it'd pull out in Spring with a blower on it!

Check us out here: www.928motorsports.com

 

Welcome aboard, Carl!

Maybe you can help clear up the most debated question here so far... How many cfm does a 928 take to get about 325 hp at redline?

Mark is estimating around 58 for a 308 from the emperical data he has from prior supercharging projects, and Scott feels it's more like 40 from the calculations he's done using data from Vortech. My numbers are somewhere in between.
Thor
www.JavelinArt.com

 

I should say, I've no statements about 928 cfm, only the 308QV which I believe apply to 328 and 308 2V engine in proportion to base hp.

Carl,
I took a look at you sight and the graghs look really good, significantly better than other engines I've seen using centrifugal superchargers as far as far as a wide usable power band.

I have some question's, when you say 6 psi, is it manifold I'm guessing? Do you have a boost curve or boost at various rpm points? And last, do you have the compressor discharge pressures at the same points?

 

I think I got an answer to my power curve shape question after finding baseline 928 graphs....928s have torque curves that peak very early then drop progressively which is a very nice match to the centrifugal compressors nothjing down low then boost that builds progressively yielding a much better looking curve than would be expected and not at all similar to 308 base curves.

I'm still interested in the boost at rpm points at the manifold and at the compressor if you have it.

 

I'll try and answer the 2 questions you have posted so far.

I thought I would receive emails from the Forum when people had posted, I know now that I do not because I have not purchased a subscription. So if you need me or an answer more quickly - try us at 920-485-0488 or [email protected]

In order for me to answer "how much boost we will need" I need to know the Discplacement of the 308 CIS motor, and the Compression Ratio too.

A lot of novices with small V-8's reach for a Vortec V-1 in S or T trim right away, and at 1200 to 1400 CFM (Cubic Feet per Minute) they have over-blown their application, and end up chasing lean fuel conditions FOREVER and they will not likely ever catch up.

Selecting the right supercharger for the application is critical - we are looking for a match to about 120% of need, and no more.

Example: on the 4.5 Liter Porsche V8 (a mear 262 cunic inches) a blower that produces 850 CFM is plenty, and more is not better. This motor produces 220 HP at the crank stock, and we raise it to 300 HP with only 6 psi, and 340 HP with 8 psi.

On the bigger Porsche 9228 engines - (1986-1995) 5.0 liter - we use a 1200 CFM blower and routinely push 400+ HP.... but they are not equipped with CIS.

Anyway - hope that helps. I cannot really state what size blower you need until someone gives me that information. There are a lot of variables - camshaft lift and duration, valve size, combustion chamber shape, exhaust system restriction, the WUR and fuel distributer, size of the fuel lines from the tank to the fuel distributer, flow of the fuel pump in LPM, and more, but I should be able to get you close.

 

i'm working on the brackets now.
carl, to answer your question, the 308 is a 3L V8 faltcrank motor. 2.9 specifically. compression ratios vary due to the way they were built. it's stated at 8.8:1 but more like 8.0:1 or less. more of the technical info is a few posts back in this thread. if you go over my calcs spreed sheet you'll find alot there as well.

the head unit i'm using is the V-9G a small blower but suited perfectly to the engine. the maps are a number posts back in this thread.

 

Scott,
Did you look at the V5-F? it's about the same dimensions as the v-9, but the specs say it's a bit lower flow and higher efficiency than the v-9, although I haven't seen the actual map....the v-9 might work out better.

I was just thinking that it might work out good if after 6800 rpm you get in to a range of decreasing efficiency where flow vs rpm starts to drop a bit as the unit approaches choke....maybe limiting the boost rise as the engines volumetric efficiency and flow drop. Just a thought. The specs say the v-5F flows about 700 scfm vs 800 for the v-9. 700ish scfm (that's at the inlet to the blower, so before the terrible efficiency heats up the air and expands to a lot more cfm) is what that first blower I tried and it had enough for the QV with efi.

You probably already look at the 5....just a though.

 

yeah i tried the V-5 first however it's in choke till 7k rpm. it's a little to small. and we'd have to spin it 5k rpms faster then the V-9. it's the same housing between the V-9 & 5 but different impellers and trim. if our air requirement is higher than i calc'd & based on your experiance then the V-9 fits better as there is still room to move right.

i'm keeping quiet on this, but as you know there is one more hurdle and it's a biggie. it rymes with 'pampers'. i'm going custom at this point. but shhhhhhh, mustn't share all the secrets going for that mass production route really opens up pandoras box!

 

Makes sence

That's certainly the nicest way to do it. I kept the stock one and drilled and pinned 3 holes in the base of the hub to center and drive the pulley.