Delivery valves
Delivery valves
full collar removal should not be a problem if you can give it enough cold air.
The problem with collar removal is the injection duration is so long there is still raw burning fuel going into the exhaust. It helps spool the turbo extremely fast but it will melt things in no time.
MUCH bigger injectors will be needed to shorten the injection duration.
I think it is VERY good to see you guys taking this little monkey wrench I tossed into the works and running with it.
Excellent idea on slotting the collar...I'm not aware of anyone doing that.
(Comparing apples:oranges moment #8356: The DI 12-valve Cummins doesn't suffer from poor idle quality until you start upgrading to larger (.120") injection lines...other than that, even your run of the mill set of shower nozzles won't affect things a whole lot.)
Beers,
Matt
GREASY_BEAST Many thanks for the tip!
GREASY_BEAST Many thanks for the tip!
Can someone explain the physics behind the delivery valve mod. From my understanding, the IP is a variable volume pump. In other words, the injection volume is controlled by the rack position. Now, I understand that increased pressures would increase the volume of fluid that bleeds past the pump barrel. I don't think that this wasted volume amounts to much. So, I have a hard time seeing how removing the delivery valve collar increases injection volume. Is there something else going on inside the pump that I am unaware of?
Bajaman Can someone explain the physics behind the delivery valve mod. From my understanding, the IP is a variable volume pump. In other words, the injection volume is controlled by the rack position. Now, I understand that increased pressures would increase the volume of fluid that bleeds past the pump barrel. I don't think that this wasted volume amounts to much. So, I have a hard time seeing how removing the delivery valve collar increases injection volume. Is there something else going on inside the pump that I am unaware of?The DV collars are one of the bottenecks. Remove the bottleneck = more fuel.
Bajaman Can someone explain the physics behind the delivery valve mod. From my understanding, the IP is a variable volume pump. In other words, the injection volume is controlled by the rack position. Now, I understand that increased pressures would increase the volume of fluid that bleeds past the pump barrel. I don't think that this wasted volume amounts to much. So, I have a hard time seeing how removing the delivery valve collar increases injection volume. Is there something else going on inside the pump that I am unaware of?The DV collars are one of the bottenecks. Remove the bottleneck = more fuel.
Its a designed restriction to delay/end injection. Remove the collar and the duration of the spray is much longer. Thats why EGT's are through the roof with the collarless DV, lots of fuel is going into the exhaust because its injecting long into the power stroke and possibly into the exhaust stroke as well.
If you want collarless DVs you would need hogged out injectors to get the spray duration much shorter.
Quote:Could it be that the fuel is injected too quickly and basically cools the combustion too much hence all the smoke.The problem is too much fuel trying to be pumped through a too small opening. Think of the injector like your thumb over a garden hose end. At low volumes its easy to hold your thumb over the end and make a spray. Turn up the flow and eventually you can't keep your thumb over the tip and the spray becomes a stream. Get a larger hose (injector), the larger diameter reduces velocity and you can once again make a spray with your thumb but flow more.
Diesel is compressible, just not by much. In a low pressure system like ours the amount is in the sub 0.1% range. Get into 30,000psi common rail systems and it can be more easy to figure.
Quote:Could it be that the fuel is injected too quickly and basically cools the combustion too much hence all the smoke.The problem is too much fuel trying to be pumped through a too small opening. Think of the injector like your thumb over a garden hose end. At low volumes its easy to hold your thumb over the end and make a spray. Turn up the flow and eventually you can't keep your thumb over the tip and the spray becomes a stream. Get a larger hose (injector), the larger diameter reduces velocity and you can once again make a spray with your thumb but flow more.
Altering the delivery valves obviously changes the injection characteristics, but I don't think that it is doing so in a favorable manner. Furthermore I don't think that it's really a bottleneck. Consider the way the delivery valve functions. A spring seats the pintle of the valve, which is forced open on each pump stroke. The shoulder of the valve restricts the flow/creates a constant pressure in the pump chamber. More flow simply means that the valve will open further. This is why I don't think you need to mess with them. The more flow, the more they open, probably with minimal changes in internal pressure. I really think that they function as a check valve AND a pressure modulator. They will flow all the pump has to offer.
Of course I could be completely wrong, therefore, I welcome anyone to disprove this thought process by actually measuring the difference in injected volumes with and without a collar.
Bajaman Of course I could be completely wrong, therefore, I welcome anyone to disprove this thought process by actually measuring the difference in injected volumes with and without a collar.
Bajaman Of course I could be completely wrong, therefore, I welcome anyone to disprove this thought process by actually measuring the difference in injected volumes with and without a collar.
Okay, You are right about increasing the injected volume. The delivery valve is has a lot of engineering involved. Most of this engineering deals with providing a SINGLE consistent pressure wave for a wide range of injection volumes. If you look at this Delivery Valve Patent, you will find gobs of information and links to older patents. All of these patents are aimed at reducing secondary injection events caused by pressure waves.
In short, yes removing the collar will inject more fuel, but this fuel will come in subsequent injection events late in the combustion cycle. So if you care nothing for fuel economy and smoke, by all means remove the collar.
No doubt the DV could be designed to operate more optimally for a certain range of injection volumes (sacrifice idle quality for larger injection volume while maintaining single injection event), but I doubt any of us know what parameters should be tweeked and in what quantities this should be done.
I guess I'll step down from my soapbox now, I'm not trying to be a prude, I just think that you are wasting effort here.
Bajaman but I doubt any of us know what parameters should be tweeked and in what quantities this should be done.
Bajaman but I doubt any of us know what parameters should be tweeked and in what quantities this should be done.
its pretty easy to deduce that the collar height has an effect on fueling volume based on modifications that other people have done for Cummins, Ford, and other engines. Reducing collar diameter would defeat the purpose of having a collar, so it can be eliminated as a parameter. Since its a cylindrical object, it has two dimensions that can be altered: radius and height. We have eliminated radius, therefore we are left with only the height as a dimension to alter.
Thickness as well. If I had access to a lathe I'd try to thin the collar by a few thousandths.
A lathe is not the proper tool to do that particular operation, unless you can get a carbide cutter small enough to make the cut. Some sort of grinder or EDM machine would probably be better. I used a makeshift grinder made using a lathe and a dremel tool, but it would be very difficult to try to make a cut with any semblance of precision using the setup I made.. Grinding the whole collar off, I was able to get them all within 3 thousandths, which is miserable. I wouldn't have been able to just take a couple thousandths off the collar, it was pretty much all or nothing. And there was no way to grind the collar's thickness, just the radius, due to the tiny amount of room available.
I had no runaway whatsoever when I ground my DVs, but then again, I did not adjust the TC to the limit, just 1/8 turn.
It will be interesting to hear what effect the partial grind has once you get the TC straightened out. My mind is totally boggled as to how you did it...
ah ok I get it now, I thought you were saying you were grinding the vertical thickness of the collar with a file, which, to me at least, seems impossible. It'll be interesting to see what a partial grind on the diameter yields.
EDIT: I wonder what the science behind DV alterations is.... someone must know...
Quote:In the course of the delivery stroke, the increasing pressure in the plunger chamber lifts the delivery-valve cone from the valve seat in the delivery-valve body. Fuel then passes through the delivery-valve holder and into the high-pressure delivery line to the nozzle. As soon as the helix of the pump plunger brings the injection process to an end, the pressure in the plunger chamber drops. The delivery-valve cone is then pressed back against the valve seat by the valve spring. This isolates the space above the pump plunger and the high-pressure side of the system from one another until the next delivery stroke.
I've started reading Bosch's Diesel In-Line Fuel-Injection Pumps and the chapter on delivery valves states the following:
Quote:In the course of the delivery stroke, the increasing pressure in the plunger chamber lifts the delivery-valve cone from the valve seat in the delivery-valve body. Fuel then passes through the delivery-valve holder and into the high-pressure delivery line to the nozzle. As soon as the helix of the pump plunger brings the injection process to an end, the pressure in the plunger chamber drops. The delivery-valve cone is then pressed back against the valve seat by the valve spring. This isolates the space above the pump plunger and the high-pressure side of the system from one another until the next delivery stroke.
Section106 So maybe the issue is the helix on the plunger? Would a thinner or smaller helix restrict enough fuel back to the fuel gallery to deliver the desired increase in fuel charge? Or maybe changing the pressure-valve spring that holds down the delivery-valve cone to one with less resistance could yield positive results?
Section106 So maybe the issue is the helix on the plunger? Would a thinner or smaller helix restrict enough fuel back to the fuel gallery to deliver the desired increase in fuel charge? Or maybe changing the pressure-valve spring that holds down the delivery-valve cone to one with less resistance could yield positive results?
winmutt Did you look at the pictures? The helix isn't the part you'd want to modify, it would be chambers at each end I would think. Still not an easy job.
Quote:Fuel delivery can be controlled by varying the effective stroke. This is acheived by means of a controlled rack which twists the pump plunger so that the pump plunger helix alters the point at which the effective delivery stroke ends and therefore the quantity of fuel delivered.
winmutt Did you look at the pictures? The helix isn't the part you'd want to modify, it would be chambers at each end I would think. Still not an easy job.
Quote:Fuel delivery can be controlled by varying the effective stroke. This is acheived by means of a controlled rack which twists the pump plunger so that the pump plunger helix alters the point at which the effective delivery stroke ends and therefore the quantity of fuel delivered.
if you have the experience, equipment, and engineering ability to modifiy the helix, then you have the experience, equipment, and engineering ability to build whatever size and/or shape elements for the MW pump you wish. In short, the helix is not modifiable.
whats it mean when an otherwise good transmission slips when I put the cut DVs in?
GREASY_BEAST whats it mean when an otherwise good transmission slips when I put the cut DVs in?You need to adjust the modulator on the tranny. Did you find a modification that keeps a even idle?
GREASY_BEAST whats it mean when an otherwise good transmission slips when I put the cut DVs in?You need to adjust the modulator on the tranny. Did you find a modification that keeps a even idle?
Hmmm... is there a procedure/method for adjusting the modulator? How much should I adjust it? Hows it done? Where is it?
I talked to Jon at JL Machine about hogging out my injectors. I plan to do this before messing with DVs again. He'll probably enlarge the radius of the holes by about 35%, and do some work on the pintle. With bigger injectors, hopefully the injection duration problem will be under control, and I can start messing with the DVs again. If I can make it so its not injecting fuel into the exhaust (thereby making waaay too high EGTs, and roasting turbochargers), then idle be damned, I'll run em. Gasser with a cam idles rough, so why shouldn't a hotrod diesel? Smoke might be a problem in the more fascist of states though... have to consider that one. Should be able to report results by the end of January. However, this transmission problem will need fixing. Hows it done?
**BTW 35% bigger hole means 1.82 times more fuel if area alone were the determining factor. But it ain't. Theres also the pintle. So it should be something like 2x more fuel.
I am also experiencing a little trouble with transmission slipping with newly found power. Here is a thread with a ton of good info about shift kits. There are also some really good pdf's in there as well. I'm excited to see the results of larger injectors and modified delivery valves! What kind of time frame are you looking at modifying you're injectors? <!-- m --><a class="postlink" href="http://schumanautomotive.com/forums/showthread.php?t=5496">http://schumanautomotive.com/forums/sho ... php?t=5496</a><!-- m -->
Lincolnlock What kind of time frame are you looking at modifying you're injectors? <!-- m --><a class="postlink" href="http://schumanautomotive.com/forums/showthread.php?t=5496">http://schumanautomotive.com/forums/sho ... php?t=5496</a><!-- m -->
Lincolnlock What kind of time frame are you looking at modifying you're injectors? <!-- m --><a class="postlink" href="http://schumanautomotive.com/forums/showthread.php?t=5496">http://schumanautomotive.com/forums/sho ... php?t=5496</a><!-- m -->
At this point its looking like a set of used Monark 265s will be the guineapigs. And I won't be doing the upgrading, Jon at JL Machine will, sometime between December and January.
interestig thread. How does the pump know to push more fuel after changing the DVs?
I thought this might be useful for reference in the future. It was posted elsewhere, and stolen, I do not have access to such beautiful diagrams.
[attachment=0]<!-- ia0 -->MWpumpelem3-1.jpg<!-- ia0 -->[/attachment]
GREASY_BEAST At this point its looking like a set of used Monark 265s will be the guineapigs. And I won't be doing the upgrading, Jon at JL Machine will, sometime between December and January.
totaldisaster How does the pump know to push more fuel after changing the DVs?
GREASY_BEAST At this point its looking like a set of used Monark 265s will be the guineapigs. And I won't be doing the upgrading, Jon at JL Machine will, sometime between December and January.
totaldisaster How does the pump know to push more fuel after changing the DVs?
The only way I can describe the logic of it, is that the fuel is compressible, if only a little tiny bit, and so is the apparatus (injector line, element, etc etc). Given these two axioms, it begins to make sense that a flow restriction inline would affect the delivered quantity of fuel. Someone slap me if this is totally off-base.
It also helps that only a miniscule amount of fuel is being pushed each time.
No. The fuel is definitely incompressible - it so so small (.00001 range). I have done the math before for liquids of petro bases. Temperature has more of an effect.
I am uncertain of the materials you are talking about, but I can get the modulus of elasticity from on line sources and do the math for those as well. Please show me which parts you are referring to, I need their shapes and sizes.
Most likely, if it is steel, and very small the change will be negligeable. Theses materials are empoyed in these situations because they are expected not to give and warp.
Right there with you on the incompressibility. I was just saying that to spark a "heads up" moment. My point is, why should removing the collar on the delivery valve, as well as making the injector hole bigger, necessarily make more fuel go into the prechamber? The only way it "works" in my mind is if there is some "flex" in the system that absorbs the injection pulse. If not, why would it matter what size the injector hole is, or whether there is a thick or thin collar on the dvalve? The amount of fuel that goes in would just be determined by the IP element's swept volume, and nothing else... do you see where I'm going with this?
GREASY_BEAST Right there with you on the incompressibility. I was just saying that to spark a "heads up" moment. My point is, why should removing the collar on the delivery valve, as well as making the injector hole bigger, necessarily make more fuel go into the prechamber? The only way it "works" in my mind is if there is some "flex" in the system that absorbs the injection pulse. If not, why would it matter what size the injector hole is, or whether there is a thick or thin collar on the dvalve? The amount of fuel that goes in would just be determined by the IP element's swept volume, and nothing else... do you see where I'm going with this?
GREASY_BEAST Right there with you on the incompressibility. I was just saying that to spark a "heads up" moment. My point is, why should removing the collar on the delivery valve, as well as making the injector hole bigger, necessarily make more fuel go into the prechamber? The only way it "works" in my mind is if there is some "flex" in the system that absorbs the injection pulse. If not, why would it matter what size the injector hole is, or whether there is a thick or thin collar on the dvalve? The amount of fuel that goes in would just be determined by the IP element's swept volume, and nothing else... do you see where I'm going with this?
I think I get what Greasy is trying to say. Basically, the DV does not determine the volume. And I would have to agree. The cam has a fixed stroke, and the bore before the DV (is that the plunger?) is not being changed. This equates to an unchanged volume. Agreed, time nad flow rate are the factors, but unless you change the volume going past the DV, modifying the DV does little to nothing.
edit: I guess I meant that its hard to imagine how modifying the DV provides more fuel considering those things.
totaldisaster I think I get what Greasy is trying to say. Basically, the DV does not determine the volume. And I would have to agree. The cam has a fixed stroke, and the bore before the DV (is that the plunger?) is not being changed. This equates to an unchanged volume. Agreed, time nad flow rate are the factors, but unless you change the volume going past the DV, modifying the DV does little to nothing.
totaldisaster I think I get what Greasy is trying to say. Basically, the DV does not determine the volume. And I would have to agree. The cam has a fixed stroke, and the bore before the DV (is that the plunger?) is not being changed. This equates to an unchanged volume. Agreed, time nad flow rate are the factors, but unless you change the volume going past the DV, modifying the DV does little to nothing.
Yes, one would think (axiomatically, that is) that modifying the collar, or the injector hole size would not result in increased fuel volume. However, this is very much not the case. I cut the collar of a set of DVs, and I can tell you for sure they flow a damn sight more fuel than the ones with a collar. I suspect a similar thing would happen with larger injector holes (because its that way with every other diesel). I was just trying to get at a better theoretical understanding of WHY!!?! A 1000FPS camera trained on the injector line, with the appropriate magnification, etc would be a way of experimentally verifying the line-flex phenomenon which I suspect is taking place. I have no such camera.
I see no other way, other than flexing of the system, that the volume of fuel injected would not be equal to the swept volume of the IP element regardless of flow restrictions. Maybe I just don't know enough about fluid dynamics.
EDIT: tomnik-- Why should the volume of fuel injected be dependent upon when the collar opens? I don't think this assumption actually correlates with what is happening... I reserve the right to be wrong.
I think I got it now.
Yesterday night I found an article (in German) describing the function of the DV.
<!-- m --><a class="postlink" href="http://www.motorlexikon.de/?I=1294&R=D">http://www.motorlexikon.de/?I=1294&R=D</a><!-- m -->
"On injection systems using long hard lines and injection pressures below 700 bar a DV type "constant volume" has to be used". (with or without restriction orifice doesn't matter).
The basic function of the DV is to keep a certain pressure in the hard line and injector gallery and to separate the pressure zones (pre and post the DV) when the effective stroke ends.
I try to describe: The plunger moves in effective stroke and the fuel pushes open the DV so the fuel can get to the injector and opens the nozzle as long as the pressure in the system is higher than the spring of the nozzle.
The flow via the DV starts only when the collar is out of the bore of the DV body.
Now the effective stroke ends. At that moment the pressure above the plunger drops. While this happens the spring of the DV forces the DV to close. On its way down the collar will separate the 2 pressure rooms when the collar enters the cylindrical part of the DV body. The nozzle is already closed and the pressure wave from the sudden dead end runs towards the DV. But (the rooms are separated) the DV still has some movement (collar in the DV body but the cone not in its seat) and takes away "some" volume (=pressure) until the cone reaches the seat. This will reduce the bouncing of the pressure wave and keeps the nozzle closed for sure.
Depending on the "stolen" volume (which is in between collar and cone the next cycle has to do more or less delivery. In the article they call it "bad or dead volume".
Now my impression: A small collar that is placed near the cone will open the flow more directly and provides a higher rest pressure in the line so the next cycle starts from a better condition and altogether more fuel is delivered. This is limited when the "stolen volume" is so small that the remaining pressure in the line (and injector) in combination with the pressure wave is high enough to open the nozzle uncontrolled.
The article also talks about individual DV to influence the delivery curves but not how this is done.
Summary: The design of the DV cone and collar (more the collar) acts like a damping device to prevent the reopening of the nozzle but therefore creates a back flow of fuel that was already delivered but comes back and so is lost for combustion.
Now such things like increased pop pressure has to be included in the future thinking.
Tom
Now THAT's interesting!! So the DV acts like a cushion absorbing the difference between the "outgoing" pressure wave from the element and the "incoming" one bouncing back from the injector.. So by that reasoning, increased pop pressure should result in more fuel pressure in the "injector room" left over from the previous cycle.. does that necessarily mean more fuel, or less fuel injected? Now that seems like a harder question to answer...
difficult to say. To build up pressure takes longer. The pressure in the line is higher. At the moment when the effective stroke ends the DV moves to close. The higher pressure in the line could close it faster (and the same amount of fuel is lost) or more fuel gets lost due to the increased pressure difference on both sides of the DV. I think in both cases less fuel is injected.
Tom
GREASY_BEAST Now THAT's interesting!! So the DV acts like a cushion absorbing the difference between the "outgoing" pressure wave from the element and the "incoming" one bouncing back from the injector.. So by that reasoning, increased pop pressure should result in more fuel pressure in the "injector room" left over from the previous cycle.. does that necessarily mean more fuel, or less fuel injected? Now that seems like a harder question to answer...
GREASY_BEAST Now THAT's interesting!! So the DV acts like a cushion absorbing the difference between the "outgoing" pressure wave from the element and the "incoming" one bouncing back from the injector.. So by that reasoning, increased pop pressure should result in more fuel pressure in the "injector room" left over from the previous cycle.. does that necessarily mean more fuel, or less fuel injected? Now that seems like a harder question to answer...
Quote:now think we have been looking at the wrong end of the scope on these pumps with barrels and plungers. Does anyone know the MAX injection pressure we can get out of the pump reliably? I think someone should start with some cut DV's and start shimming their injectors until the pump will no longer inject. You can pull one injector from the engine and run it live on the other side of the pump. We could set a target of 50% the distance between 135bar and the max.
Quote:now think we have been looking at the wrong end of the scope on these pumps with barrels and plungers. Does anyone know the MAX injection pressure we can get out of the pump reliably? I think someone should start with some cut DV's and start shimming their injectors until the pump will no longer inject. You can pull one injector from the engine and run it live on the other side of the pump. We could set a target of 50% the distance between 135bar and the max.
You guys must have seen this already:
<!-- m --><a class="postlink" href="http://www.w124performance.com/docs/Bosch/PreTech_4.3_MW-Pump.pdf">http://www.w124performance.com/docs/Bos ... W-Pump.pdf</a><!-- m -->
one of the members on M-shop organized this site. Hope this pump doesn't differ too much.
According to this guide within the first few pages. Adjustment of the rack will change stroke and deliver more fuel. Imagine the helix advanced reaching the port sooner. - did I miss this before?
Again - the steel flexing is not a major part of this equation. Give me the dimensions of the parts you are considering and I will do the math to attempt to prove it.
why would they design brake lines to swell? how are these different?
oh. Just thought of something.
with the collar removed the return time is slower for the DV to fully close - correct? I think you may be sucking fuel in that would otherwise be flowing out the injector return lines.
The plunger would be partially filled before the helix reaches the port and any fuel added could be considered your 'normal stroke'.
maybe?