|
Indexing
Originally posted by Don M No, extrude honing does not really change anything on the spray pattern. The impact remains roughly the same. The drawback to EH is the internal seat destruction.... if it is not done properly. Erosion of the metal you need in the seat is washed away with aggressive/improper honing. Recall the nozzles are lightly honed from Bosch to reduce the flow tolerance between them all. I stress lightly. Not all EH'ing is created equal. We have some 125 HP EH nozzles that never pop or bang. Most others cant break the 80 HP mark without the poping at high rpm. At last check, I have a guy with just over 90,000 miles on some honed 90 HP injectors. It can be safe, if its done properly. EDM just gives a total flexibility to make whatever you like. Hole size, number of, angle of the cone, etc. This can eliminate honing altogether and get the flow rate you want. Don~ Does indexing play a part in selecting the proper spray angle? I've been following your posts for some time as have many others,....how close are you on EDMs for the 600? |
Yes indexing is important. The spray angle is paramount. When we try a new design we move the spray angle, change the hole sizes, the number of, etc. Knowing the injection angle numbers ( timing ), injected fuel quantity, the discharge curve ( rate of injection ) injection pressure are only a small portion of the things considered.
We can then run everything in advance through computational software and mathematical models. These models give us a leg up before we ever pop a single hole. Some of the best sounding ideas turn out to be the worse in reality. We have run simulations of over 200 "ideas" Most of them go by the wayside. Most of the software simulations can take hours to run because I get limited time on the big computers. I have left models running overnight to simulate a single combustion cycle. Literally hundreds of hours goes into one design in simulation. If I had some additional time on the big hardware or a larger cluster of boxes strung together it would be faster. We are not bound by the NOx and soot requirments of the OEM's, but you would be surprised just how close a low soot and Nox injector is to a performance injector. The OEM are trying to ring out as much performance per component as well. Power density per litre is way up in modern Diesels. The EDM's for the 600 have been running in the public for a few months. This was after extensive design, machining and power testing. I cant stress how much work goes into it. Even if we feel we have the best thing going and the product works well....I still insist on long term data and testing. Heck I have had a handful of prototype pressure box testers since April. We have still yet to sell a single one and they are considered a final product now. If everything goes good..in the next few months we will have them ready for the 600. The 03-04 will be ready right after the holidays!! A few guys have them now, a few guys have them ordered and are waiting. Don~ |
LOL.... Don, need to borrow some computing power?
I have some watercooled computers that love to crunch numbers 2 AthlonXP2400 2 Athlon 2800 and 1 laptop3.6 All Running F@H |
Indexing
Originally posted by Don M Yes indexing is important. The spray angle is paramount. When we try a new design we move the spray angle, change the hole sizes, the number of, etc. Knowing the injection angle numbers ( timing ), injected fuel quantity, the discharge curve ( rate of injection ) injection pressure are only a small portion of the things considered. We can then run everything in advance through computational software and mathematical models. These models give us a leg up before we ever pop a single hole. Some of the best sounding ideas turn out to be the worse in reality. We have run simulations of over 200 "ideas" Most of them go by the wayside. Most of the software simulations can take hours to run because I get limited time on the big computers. I have left models running overnight to simulate a single combustion cycle. Literally hundreds of hours goes into one design in simulation. If I had some additional time on the big hardware or a larger cluster of boxes strung together it would be faster. We are not bound by the NOx and soot requirments of the OEM's, but you would be surprised just how close a low soot and Nox injector is to a performance injector. The OEM are trying to ring out as much performance per component as well. Power density per litre is way up in modern Diesels. The EDM's for the 600 have been running in the public for a few months. This was after extensive design, machining and power testing. I cant stress how much work goes into it. Even if we feel we have the best thing going and the product works well....I still insist on long term data and testing. Heck I have had a handful of prototype pressure box testers since April. We have still yet to sell a single one and they are considered a final product now. If everything goes good..in the next few months we will have them ready for the 600. The 03-04 will be ready right after the holidays!! A few guys have them now, a few guys have them ordered and are waiting. Don~ Is the indexing done via shimming and if so then this would have to be done at installation or do you do it via some other methodology to provide indexing? When will you be willing to take orders for 600 injectors? |
GJ,
The source codes are FORTRAN and C. Most are compiled to run under UNIX. Silicon Graphics, Cray, Sun, and HP. I like the time on the Cray. The time to compute is cut way down. A fast Sun workstation with minimal grid points takes hours in 2-D. Comparing; the Cray will crunch a 3-D, 25,000 grid point simulation in minutes with extensive complex chemistry and many spray parcels. Its hard to go back and try that on a UNIX/LINUX box workstation. It takes hours longer. All you can do is let the cooling fans sing while you try to sleep at night. Some of the routines I have run kicked out gigabytes of data!!!! The transition to CFD for sprays was easy for me coming from Fire Protection. We ran many simulation codes for everything from smoke models to where nuclear fallout would land. Of course the CFD is much more complex and the number of hours is higher by a large factor. The longest part in terms of time and my limited patience with some things was the grid generation for three different engines and piston designs. 12, 24 and ISBe engines may all seem the same but an offset bowl in a 12 valve style grid wont work with a centered bowl design in the 24 valve. In the future I planned on having available a small amount of code for everyone to use. Something simple to run under Windoz where everyone could see what their injectors were doing in-cylinder. Something pre-compiled and ready to run with minimal hardware needed. Too bad I cant afford to buy or have a way to get a Cray in my livingroom. LMAO! One day we will all have the computing power of a Cray in our laptops. You know the Saturn 5 rocket controls had less computing power than a cheesy 133 Mhz laptop does noow. LOL!! O-OF-H, When you said indexing...I thought you meant spray hole indexing in regards to where the holes were located in relation to the injector position. You cant change much on shim height. The connector tubes to injector interface gets out of whack too much. If you need a change its easier to just raise or lower the spray hole in the nozzle. No 600 stuff for a few months at least. Don~ |
Don.... no cooling fans here!
Nothing but a heater core for a radiator (Ford SD) and a magdrive 350GPH Pump and a 5 gallon resevoir. Glorious silence.... although only 2 systems are running a parallel processing unix kernel. Too underpowered for what you need.:D |
Sweet! What do you run on it? CAE?
RAM is the main thing needed. gigs Don~ |
between the 2 systems 2 gigs of registered ddr333
CAE? |
Computer Aided Engineering.
Most normal guys dont have these kinds of toys for gaming. LOL Must be RAMMED for something? Science? Don~ |
Yes. A Cancer Research Program From Stanford University.
Do a Google search for 'folding at home' Really stresses a system. And as Overclocked as these rigs are, they have to be watercooled. Running 24% faster than designed.... total thermal load for the 4 systems just on processor heat is 577 Watts. Water temp is a warm 36C on a 17C ambient room. Helps to heat the house though! |
Ahhh, distributed. Now I get it. The power of over 1,000,000 processors cranking is insane. Cancer research makes more sense than climate models, IMO.
Heats the house. LOL Very cool Juice. I want to hook in myself now. I dont have the power you have to donate, but I have a few flops per year to share. Don~ |
Too bad we couldn't hook some injector modeling into that process... results in minutes instead of hours.
So ... back to topic, do you think the DMX needs more holes .. smaller than stock for a perf injector? a 25% reduction in hole diameter and 4 more holes? |
I dont know what is best. The variables between the DMAX and the Cummins alone are enough to cause me reservation. Im behind on the DMAX.
Don~ |
Originally posted by Don M The source codes are FORTRAN and C. Most are compiled to run under UNIX. Silicon Graphics, Cray, Sun, and HP. I like the time on the Cray. The time to compute is cut way down. A fast Sun workstation with minimal grid points takes hours in 2-D. Comparing; the Cray will crunch a 3-D, 25,000 grid point simulation in minutes with extensive complex chemistry and many spray parcels. Its hard to go back and try that on a UNIX/LINUX box workstation. It takes hours longer. All you can do is let the cooling fans sing while you try to sleep at night. Some of the routines I have run kicked out gigabytes of data!!!! Andy |
Don M,
You had it right, I was talking about indexing relative to injector postition,....but I was talking about spray angles relative to the valves position once the injector is installed and to insure that those angles stay or keep that same relativity, then it seems that they (injectors) would need to be preinstalled into position to insure that the angles stay as designed relative to the valves and I know indexing can accommodate that. Right?
|
| All times are GMT -5. The time now is 08:21 PM. |
© 2024 MH Sub I, LLC dba Internet Brands