DuAl FeDe fewyl RaILS 4 da WIn?
06-03-2008, 02:50 PM
#41
Elite Member
Thread Starter
iTrader: (3)
Join Date: Aug 2006
Location: San Diego, CA
Posts: 3,047
Total Cats: 12
Actually Abe, the fuel rails on the 99-00 do feed through the back. On the front is a pressure dampener. If you examine it, youll see that the dampener on the front isnt really connected to anything. Instead of cylinder 4 running really hot, (like most cars) our cars have cylinder 1 burning up.
OMG, you're right. There's a inlet nipple on the front, but it just runs to the back before feeding anything.
Well, converting it to "dual feed" would just be a matter of swapping the regulator for a nipple and putting a Y in the line.
Think that's a good way to go? Sure is easy.
Yes, on the early cars it gets done all the time. I think on our later ones it could be done, as well.
Reply
0
0
06-03-2008, 02:55 PM
#42
Boost Pope
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,026
Total Cats: 6,592
Hmmmmm.
First off, if someone wants to donate a fuel rail w/ FPR (any NA), a fuel pump, and four high flow injectors (must be new or recently flowtested) then I'd be happy to do the experiment and videotape it. We don't need pressure gauges, just four graduated cylinders and a stopwatch.
I came in here expecting to say "Bollocks", however I've done some cipherin' and now I'm not so sure. First, let us refer to the following hypothetical:
We have a rail with an inlet, an outlet, two injectors, and a restriction in the middle. It is true that liquids are non-compressible and that in an open system the pressure of a liquid will be constant in all areas. However where flow and restriction are present, there can be pressure differentials.
In the above diagram, assume that the flow coming in at point A is not significantly greater than the total system requirement (B + C) and that therefore the flow out of D will be small. Secondarily, assume that the restriction is quite severe. In this case, there will in fact be a pressure differential between B and C, and the flow at B will be greater than the flow at C.
But I figured "Hey, the truth of the matter is quite different, isn't it?" I'm not so sure.
Here's a more realistic system, which I will refer to from here on:
Judging by the chart posted by Braineack I deem that the performance of a "155lph" pump (comparable to stock) at 50 PSI is ~ 35GPH. This translates to 2,200 cc/min.
Using the RC Engineering calculator, I have determined that to produce 300 crank HP assuming .6 BSFC requires appx. 475 cc/min of fuel per injector, or 1,900 cc/min total. This is achievable with a 550cc rated injector running ~85% DC.
Ok, so assume that the flow coming into A is 2,200 cc/min, and that the FPR is working to produce 50PSI at E. Assuming that 1,900 cc/min is being injected, the FPR is bypassing 300 cc/min back into the tank. IOW, the amount of fuel being bypassed is less than the amount of fuel coming out of any one injector.
This actually surprises me a great deal. It also makes me wonder whether there may actually be some validity to the dual feed concept.
The question really becomes one of how restrictive the rail itself is at 2,200 cc/min of flow. I don't even know what the ID of the stock rail is, nor how I'd go about calculating the drop across it if I did. If it is restrictive, then what you'd see, technically, is not a reduction of pressure at points D or E, but rather an increase in pressure at point A, since the pressure reference is the FPR. So technically injectors 1 and 2 would be running rich, though by the time you even everything out and achieve target AFR overall, the net effect would be for 3 and 4 to be running slightly lean.
The big question, of course, is by how much and does it really matter?
My gut feeling however is that the rail is not restrictive. Just think about the diameter of the hole(s) in the bottom of the injector. They're pretty damn tiny, and yet they're flowing 500+ cc/min. Multiply by four, and you've still got a tiny hole. Compare that to the I.D. of the rail. And of course, the whole volume of fuel isn't flowing through any one portion of the rail. 1/4 of the useful fuel goes off to the right to feed injector 1, and only 3/4 of the useful fuel flows past injector 2.
AFRs are pretty much a crap-shoot anyway, and most of us tend to tune on the rich side for safety. That's pretty much my final answer at the moment- that during the one condition where flow and restriction do become meaningful, we're already operating the engine in such a rich condition that a small imbalance is unlikely to be meaningful.
Still, I'm open to donations.
First off, if someone wants to donate a fuel rail w/ FPR (any NA), a fuel pump, and four high flow injectors (must be new or recently flowtested) then I'd be happy to do the experiment and videotape it. We don't need pressure gauges, just four graduated cylinders and a stopwatch.
I came in here expecting to say "Bollocks", however I've done some cipherin' and now I'm not so sure. First, let us refer to the following hypothetical:
We have a rail with an inlet, an outlet, two injectors, and a restriction in the middle. It is true that liquids are non-compressible and that in an open system the pressure of a liquid will be constant in all areas. However where flow and restriction are present, there can be pressure differentials.
In the above diagram, assume that the flow coming in at point A is not significantly greater than the total system requirement (B + C) and that therefore the flow out of D will be small. Secondarily, assume that the restriction is quite severe. In this case, there will in fact be a pressure differential between B and C, and the flow at B will be greater than the flow at C.
But I figured "Hey, the truth of the matter is quite different, isn't it?" I'm not so sure.
Here's a more realistic system, which I will refer to from here on:
Judging by the chart posted by Braineack I deem that the performance of a "155lph" pump (comparable to stock) at 50 PSI is ~ 35GPH. This translates to 2,200 cc/min.
Using the RC Engineering calculator, I have determined that to produce 300 crank HP assuming .6 BSFC requires appx. 475 cc/min of fuel per injector, or 1,900 cc/min total. This is achievable with a 550cc rated injector running ~85% DC.
Ok, so assume that the flow coming into A is 2,200 cc/min, and that the FPR is working to produce 50PSI at E. Assuming that 1,900 cc/min is being injected, the FPR is bypassing 300 cc/min back into the tank. IOW, the amount of fuel being bypassed is less than the amount of fuel coming out of any one injector.
This actually surprises me a great deal. It also makes me wonder whether there may actually be some validity to the dual feed concept.
The question really becomes one of how restrictive the rail itself is at 2,200 cc/min of flow. I don't even know what the ID of the stock rail is, nor how I'd go about calculating the drop across it if I did. If it is restrictive, then what you'd see, technically, is not a reduction of pressure at points D or E, but rather an increase in pressure at point A, since the pressure reference is the FPR. So technically injectors 1 and 2 would be running rich, though by the time you even everything out and achieve target AFR overall, the net effect would be for 3 and 4 to be running slightly lean.
The big question, of course, is by how much and does it really matter?
My gut feeling however is that the rail is not restrictive. Just think about the diameter of the hole(s) in the bottom of the injector. They're pretty damn tiny, and yet they're flowing 500+ cc/min. Multiply by four, and you've still got a tiny hole. Compare that to the I.D. of the rail. And of course, the whole volume of fuel isn't flowing through any one portion of the rail. 1/4 of the useful fuel goes off to the right to feed injector 1, and only 3/4 of the useful fuel flows past injector 2.
AFRs are pretty much a crap-shoot anyway, and most of us tend to tune on the rich side for safety. That's pretty much my final answer at the moment- that during the one condition where flow and restriction do become meaningful, we're already operating the engine in such a rich condition that a small imbalance is unlikely to be meaningful.
Still, I'm open to donations.
Reply
0
0
06-03-2008, 03:35 PM
#43
I think an even smarter way to measure if a dual feed is affective or not is to measure pressure at each of the four points where there is an injector. If the pressure there is over 43.5 and I will put money on the fact that it always is, dual feed wont make a difference.
Something tells me the graduated cylinder test wont work too well.
Something tells me the graduated cylinder test wont work too well.
Reply
0
0
06-03-2008, 03:52 PM
#45
Boost Pope
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,026
Total Cats: 6,592
Merely measuring the pressure at each injector location will tell us nothing unless the engine is actually running at full load, or at least, unless the injectors have been run up to their effective maximum duty-cycle. We need to simulate the maximum flow through the rail, as that's what creates pressure drop. Performing that test without a dyno would require removing the rail from the engine anyway.
And of course, we're not looking for pressure to simply be above any particular pressure, but rather that the pressure at each injector is the same. Using four pressure gauges would work, but I think squirting the injectors into a cup would just be easier to do, assuming you have four injectors whose flow rates are known to be identical.
And of course, we're not looking for pressure to simply be above any particular pressure, but rather that the pressure at each injector is the same. Using four pressure gauges would work, but I think squirting the injectors into a cup would just be easier to do, assuming you have four injectors whose flow rates are known to be identical.
Reply
0
0
06-03-2008, 04:33 PM
#48
Merely measuring the pressure at each injector location will tell us nothing unless the engine is actually running at full load, or at least, unless the injectors have been run up to their effective maximum duty-cycle. We need to simulate the maximum flow through the rail, as that's what creates pressure drop. Performing that test without a dyno would require removing the rail from the engine anyway.
And of course, we're not looking for pressure to simply be above any particular pressure, but rather that the pressure at each injector is the same. Using four pressure gauges would work, but I think squirting the injectors into a cup would just be easier to do, assuming you have four injectors whose flow rates are known to be identical.
And of course, we're not looking for pressure to simply be above any particular pressure, but rather that the pressure at each injector is the same. Using four pressure gauges would work, but I think squirting the injectors into a cup would just be easier to do, assuming you have four injectors whose flow rates are known to be identical.
As for squirting injectors into a cup.
If you want to do it that way, you first have to get a baseline of exactly how much each injector will produce wit its given flow(Because they all vary slightly). So that means very simple capping off 3 ports on the rail and firing just one injector for a specified amount of time. Repeating for all four to get a baseline for all the injectors. Then running all of them and seeing what comes out and also taking into account the variance in their flow.
and youre sure thats the reason how?
Have you considered that perhaps the hot fluid circling around the first cylinder from the mixing manifold might be increasing temps at cylinder one?
Thought I had for example.
Reply
0
0
06-03-2008, 04:58 PM
#49
It is official... $99 Billet Rails.. Not a Rail stock/welded tabs rail! See threads below..
https://www.miataturbo.net/forum/showthread.php?t=21894
https://www.miataturbo.net/forum/showthread.php?t=21895
I told you all it was coming!
https://www.miataturbo.net/forum/showthread.php?t=21894
https://www.miataturbo.net/forum/showthread.php?t=21895
I told you all it was coming!
Reply
0
0
06-03-2008, 05:13 PM
#50
Elite Member
Thread Starter
iTrader: (3)
Join Date: Aug 2006
Location: San Diego, CA
Posts: 3,047
Total Cats: 12
A while ago, I had several injectors, old crusty, dirty, chipped, nasty "310 cc" injectors from a high mileage junk yard volvo.
I took the feed line from my fuel rail, hose clamped it to the injector, put 12V on the injector, held it for 30 seconds, and got the gas in a cup.
All done by hand. Then I weighed it, and they were within 5 or 8% or so.
So it's not that hard to run the test, hand timed and everything.
The reason I want to use the megasquirt for it is that I can simulate the order they fire in, if there ARE significant pressure waves (and I assume there are some or my car wouldn't have come with several pressure dampers on it), this might show.
I'm not against the graduated cyl test. I might put a piece of tubing on each injector to make sure you catch all the fuel, and have it feed a cup with a good sensitive balance instead of trying to read volume, but either way would work. Evaporation is my biggest concern.
If you measured pressure, where would you measure it? Sounds complicated, and I want to take into account a lot about the injectors. If it's low pressure just as it opens, it won't show on the average pressure at the guage. The flow of the injector may mess up your reading. We don't care about pressures, we care about fuel flow!
I obviously am a bit interested in my non-regulated rail. :-) But the same test set up will be easy, last time I did it all off my car, letting it's supply to all the work. That gives me realistic conditions, I feel.
Christ, that's close! Well, I don't trust that RC Engineering thing so much, I know people making lots more power than they quote.
Here's another thing, at 85% duty cycle (forgetting the time to get up to flow and down from it) you've got another 30% higher rates at the injector as the pulse. Still, 550cc/min / 60sec/min * 0.01 ms cycle time at 6,000 rpm * 0.8 duty cycle means 73 microliters of fuel in a single injection. That's a cube less than half a cm on a side. Hmmm, ok, that IS about the dead volume around the injector inlet, so.... ok, so I don't know what that means. :-)
I'm with you here, wow that's not a lot of margine. True, you're not using that often, or for long, but it's something to be aware of. I know folks making 350 whp on 450cc's of injector, but it's ballpark. I don't think many of the 225 hp miatas are hitting this limit.
It's way better than you think, restriction to fluid flow goes as the 4th power of the diameter. So the injector holes, let's pretend they are 1/10th the diameter even though it's much less, have 10^4 or 10,000 times the restriction. The easy test here is take a piece of tube, pump fuel up into it, drill 4 holes, each 0.010", and see if there's still pressure at the far end. I bet so.
Yes, and no. Much like timing, and everything else, the reason the modern engines have such high specific power is because you have cut the margin of safety down so much. If you can get the injectors to be dead on equal, where before they were 5% off, then you can get your AFR's 5% closer to the edge.
I still think it's much more important on timing... to be dead on.
It would be nice, but I think we can accomadate that by knowing what they are by putting them in the same places in different rails, and/or doing a quick flow test before the test.
I have an FM rail with a feed on each end (technically, an inlet and an outlet) and some injectors that fit it, but it's for my piggy, so the only really interesting thing we could do with it is feed it from one end with the other capped verses feeding it at both ends, and see what the difference is.
I took the feed line from my fuel rail, hose clamped it to the injector, put 12V on the injector, held it for 30 seconds, and got the gas in a cup.
All done by hand. Then I weighed it, and they were within 5 or 8% or so.
So it's not that hard to run the test, hand timed and everything.
The reason I want to use the megasquirt for it is that I can simulate the order they fire in, if there ARE significant pressure waves (and I assume there are some or my car wouldn't have come with several pressure dampers on it), this might show.
I think an even smarter way to measure if a dual feed is affective or not is to measure pressure at each of the four points where there is an injector. If the pressure there is over 43.5 and I will put money on the fact that it always is, dual feed wont make a difference.
Something tells me the graduated cylinder test wont work too well.
Something tells me the graduated cylinder test wont work too well.
If you measured pressure, where would you measure it? Sounds complicated, and I want to take into account a lot about the injectors. If it's low pressure just as it opens, it won't show on the average pressure at the guage. The flow of the injector may mess up your reading. We don't care about pressures, we care about fuel flow!
Hmmmmm.
First off, if someone wants to donate a fuel rail w/ FPR (any NA), a fuel pump, and four high flow injectors (must be new or recently flowtested) then I'd be happy to do the experiment and videotape it. We don't need pressure gauges, just four graduated cylinders and a stopwatch.
First off, if someone wants to donate a fuel rail w/ FPR (any NA), a fuel pump, and four high flow injectors (must be new or recently flowtested) then I'd be happy to do the experiment and videotape it. We don't need pressure gauges, just four graduated cylinders and a stopwatch.
Judging by the chart posted by Braineack I deem that the performance of a "155lph" pump (comparable to stock) at 50 PSI is ~ 35GPH. This translates to 2,200 cc/min.
Here's another thing, at 85% duty cycle (forgetting the time to get up to flow and down from it) you've got another 30% higher rates at the injector as the pulse. Still, 550cc/min / 60sec/min * 0.01 ms cycle time at 6,000 rpm * 0.8 duty cycle means 73 microliters of fuel in a single injection. That's a cube less than half a cm on a side. Hmmm, ok, that IS about the dead volume around the injector inlet, so.... ok, so I don't know what that means. :-)
Ok, so assume that the flow coming into A is 2,200 cc/min, and that the FPR is working to produce 50PSI at E. Assuming that 1,900 cc/min is being injected, the FPR is bypassing 300 cc/min back into the tank. IOW, the amount of fuel being bypassed is less than the amount of fuel coming out of any one injector.
This actually surprises me a great deal. It also makes me wonder whether there may actually be some validity to the dual feed concept.
This actually surprises me a great deal. It also makes me wonder whether there may actually be some validity to the dual feed concept.
My gut feeling however is that the rail is not restrictive. Just think about the diameter of the hole(s) in the bottom of the injector. They're pretty damn tiny, and yet they're flowing 500+ cc/min. Multiply by four, and you've still got a tiny hole. Compare that to the I.D. of the rail. And of course, the whole volume of fuel isn't flowing through any one portion of the rail. 1/4 of the useful fuel goes off to the right to feed injector 1, and only 3/4 of the useful fuel flows past injector 2.
AFRs are pretty much a crap-shoot anyway, and most of us tend to tune on the rich side for safety. That's pretty much my final answer at the moment- that during the one condition where flow and restriction do become meaningful, we're already operating the engine in such a rich condition that a small imbalance is unlikely to be meaningful.
I still think it's much more important on timing... to be dead on.
I have an FM rail with a feed on each end (technically, an inlet and an outlet) and some injectors that fit it, but it's for my piggy, so the only really interesting thing we could do with it is feed it from one end with the other capped verses feeding it at both ends, and see what the difference is.
Reply
0
0
06-03-2008, 06:36 PM
#51
I'm excited....subscribing for our very own Miata Mythbusters experiment...
I hope you guys put an end to this one way or another so its one more question answered...Please dont let this be a 15pg thread going back and forth for nothing...
I expect to see another JoePerez video real soon!!!
I hope you guys put an end to this one way or another so its one more question answered...Please dont let this be a 15pg thread going back and forth for nothing...
I expect to see another JoePerez video real soon!!!
Reply
0
0
06-03-2008, 07:23 PM
#55
Elite Member
Thread Starter
iTrader: (3)
Join Date: Aug 2006
Location: San Diego, CA
Posts: 3,047
Total Cats: 12
Mach,
If he doens't, I do - but odds are we'd do it together anyway. :-)
I've got injectors, though I guess throwing in more would let us flow match them well before starting. Do you have the regulator on there?
Now I'll just have to come up with an appropriate hat to wear during such a test. And maybe a dual feed 99 rail.
----------------
Aside:
I know people are big on the dampers, I'm not sure if they are for damping fluctuations from the pump or fluctiations from the injectors (I suspect the later). If so, it would seem just having both ends of the rail open, and going to (relatively) expandable hose would greatly damp such fluctuations... Such that you could just take a stock '99 rail, remove the damper, put on a barb, and run a second line to it, which would both damp pulses and even up the distribution.
I've got my 550's, I have a friend with some 750's, I'll see if I can borrow them. I hope they fit the rail.
If he doens't, I do - but odds are we'd do it together anyway. :-)
I've got injectors, though I guess throwing in more would let us flow match them well before starting. Do you have the regulator on there?
Now I'll just have to come up with an appropriate hat to wear during such a test. And maybe a dual feed 99 rail.
----------------
Aside:
I know people are big on the dampers, I'm not sure if they are for damping fluctuations from the pump or fluctiations from the injectors (I suspect the later). If so, it would seem just having both ends of the rail open, and going to (relatively) expandable hose would greatly damp such fluctuations... Such that you could just take a stock '99 rail, remove the damper, put on a barb, and run a second line to it, which would both damp pulses and even up the distribution.
I've got my 550's, I have a friend with some 750's, I'll see if I can borrow them. I hope they fit the rail.
Reply
0
0
06-03-2008, 08:03 PM
#57
Elite Member
Thread Starter
iTrader: (3)
Join Date: Aug 2006
Location: San Diego, CA
Posts: 3,047
Total Cats: 12
I'd certainly be in. Just up to Joe to chime in if he wants the pump or not. I think I have some 55 gph pump on my bike I'm not using. :-) Low pressure though.
I'd be pretty tempted to try feeding it from both ends by making an adapter. It doesn't really look very different from the NB one, functionally it's just a square rail with some holes along it.
Edit: Any chance I could talk you down from $0? Say, maybe, gimmie $15 to take it off your hands?
I'd be pretty tempted to try feeding it from both ends by making an adapter. It doesn't really look very different from the NB one, functionally it's just a square rail with some holes along it.
Edit: Any chance I could talk you down from $0? Say, maybe, gimmie $15 to take it off your hands?
Reply
0
0
06-03-2008, 08:15 PM
#58
I'm not against the graduated cyl test. I might put a piece of tubing on each injector to make sure you catch all the fuel, and have it feed a cup with a good sensitive balance instead of trying to read volume, but either way would work. Evaporation is my biggest concern.
If you measured pressure, where would you measure it? Sounds complicated, and I want to take into account a lot about the injectors. If it's low pressure just as it opens, it won't show on the average pressure at the guage. The flow of the injector may mess up your reading. We don't care about pressures, we care about fuel flow!
You are right about the pressure 100%, its complicated and its not the results we are looking for.
Reply
0
0
06-03-2008, 08:43 PM
#59
So, short of tests, it sounds like the conclusions that the average person should take away thus far are:
1) Other than bling factor and 'peace of mind', stock rail should not be in question until approaching or exceeding ~250whp.
2) It's likely that you'll start hitting the limits of the stock fuel pump around this time.
3) A non-scientific evaluation of "this hole is a lot bigger than this hole" suggests the injector is more of a bottle-neck than the rail, assuming adequate flow from the pump
If a test is to be performed, I can offer a set of 305cc Supra injectors. (They're arriving shortly, then off to Injector Rehab for cleaning..) Seems the desirable testing would be stock, mid-ground (e.g. the 305's), and RC 550's.
Just because it was chosen to be SUFFICIENT for 102rwhp, doesn't mean it wont be sufficient for 5-10 times as much. After all, the whole point is we're taking an entire car that was "designed" for one thing, and making it do much much more, but only by upgrading the necessary parts that were insufficient from the factory.
To that end, keep in mind, the factory doesn't design all parts from scratch for each car... they re-use parts when available, and when they're making a new part, they usually try to make it as cheaply as possible (a better spec may happen to be cheaper), and to be able to re-use that part in the future. E.g. look at all the cross-over between the MX-5 and the RX-7. That's not a coincidence.
1) Other than bling factor and 'peace of mind', stock rail should not be in question until approaching or exceeding ~250whp.
2) It's likely that you'll start hitting the limits of the stock fuel pump around this time.
3) A non-scientific evaluation of "this hole is a lot bigger than this hole" suggests the injector is more of a bottle-neck than the rail, assuming adequate flow from the pump
If a test is to be performed, I can offer a set of 305cc Supra injectors. (They're arriving shortly, then off to Injector Rehab for cleaning..) Seems the desirable testing would be stock, mid-ground (e.g. the 305's), and RC 550's.
To that end, keep in mind, the factory doesn't design all parts from scratch for each car... they re-use parts when available, and when they're making a new part, they usually try to make it as cheaply as possible (a better spec may happen to be cheaper), and to be able to re-use that part in the future. E.g. look at all the cross-over between the MX-5 and the RX-7. That's not a coincidence.
Reply
0
0
