Computer simulation of supersonic jet outflow into a prechamber with different values of initial pressure.


The objective of this study was in numerical analysis of M=2 axi-symmetric jet interaction with rigid inclined surface located in a prechamber below a nozzle exit. The shape of the jet and flow field parameters for different initial pressures in the prechamber (from normal 105 Pa down nearly to a pure vacuum – 0.133 Pa) were analyzed.

The problem was considered in steady state. At the entrance of the nozzle a total pressure of 2 MPa and temperature of 600 K were specified. The exit diameter of the nozzle is equal to 10 mm. The initial temperature in the prechamber and wall temperature was equal to 300 K.

At the initial state 2D axi-symmetryc flow inside the nozzle was calculated. Fig 1 shows distribution of Mach number and static pressure in convergent-divergent nozzle. A sound line in the throat of the nozzle and consequent acceleration of the flow are clearly seen. Zones of generation and interaction of shock waves and rarefaction waves are recognized in supersonic flow.



3D simulations of supersonic outflow out of the nozzle were performed in a domain that included the nozzle, the prechamber and the inclined plate. Fig. 2 represents the geometry of half-domain with vertical symmetry plane. A hexa mesh with 800 cells was generated in the domain. Fig. 3 shows a fragment of the mesh in a vertical symmetry plain.



A Reynolds averaged Navier-Stokes equations (RANS) together with standard 2 parametric k-epsilon turbulence model, that is typically used for jet flows, are solved numerically. Fig. 4-18 shows in a vertical symmetry plain calculated distributions of velocity field, streamlines, Mach number, static and total pressures, temperatures and energies of turbulent pulsation for 4 different values of initial pressure in the prechamber: 105, 0.5x105 , 104, and 0.133 Pa. The shape of the jet and distribution of flow parameters dramatically changes with decrease of the pressure in the prechamber.

As one can see from Fig. 4-10 for the prechamber pressures of 105 and 0.5 105 Pa supersonic axi-symmetric jet initially widens and accelerates (Mach number in the first “barrel?” reaches 3.9 and 5.2, correspondingly), and then external pressure starts to narrow the jet forming a structure with “barrels” resulting in widening and narrowing the jet shape. Up to 3-4 “barrels” can be observed within calculation domain. For prechamber pressure of 105 Pa (left figures) the width of the jet is smaller than in the case of 0.5 105 Pa pressure. At the position where the jet interacts with the plate it curves and attaches to the plate as a result of Coanda effect.







For lower prechamber pressures (104 and 0.133 Pa, Fig. 11-18) the shape of the jet is strongly different. Tenfold decrease in pressure in the prechamber leads to development of the jet with only one “barrel” (left column of figures). For lower pressure (0.133 Pa) the jet expands rapidly (right column of figures) straight after leaving the nozzle and terminates with inclined compression shock wave, that divides high-velocity supersonic flow inside the jet and much slow sub/supersonic flow in the vicinity of inclined plain.

Computer simulation of supersonic outflow of viscous gas jet into a prechamber with different initial pressure and its interaction with a rigid inclined wall reveals significant changes of the jet shape and flow characteristics depending on the prechamber pressure. The shape changes from a jet with “barrels” for normal atmospheric pressure to a rapid expanding jet after nozzle exit for the case of negligibly small back pressure (0.133 Pa).
Mach number distribution

Why are you modeling using such low initial pressures in the PC? To gain a sense of what is happening with the flow?

I came across it during some recent research. The math is over my head, but I can see some interesting things in the images. I am betting that the results are scaleable to some extent.

My main interest is the affect of a flow pattern through an orifice when the differential pressure raises, and then is reduced. I have done two stages of PC mods with initial testing in an "environment", in the shop, and the latest testing installed in the engine.

My free time has been limited, so the tests were "quick and dirty", just to indicate if this direction is moving in the desired direction. The documentation of the testing environments/conditions, processes, and results are less than "scientific", and would surly start an argument as to the validity of the tests and any perceived results. I have seen it happen before. I will wait until I have the 1st gen prototypes installed before I post any performance results.

The next step is to make some new lower halves for some angled injector PCs so they will fit in the 167a and early 616 engines. 99% of the modifications I am doing are to the lower half of the PC.

Unfortunately the angled PCs are pricey, and I am going to use Inconel to make the new lower sections which is pricey again. Additionally, I will need to have some EDMing done as well as getting the lower injector housings swapped to fit the angled PCs. Even further, considering the big picture, (larger Elements), I want to get some extrude honed nozzles put in the injectors when the lower halves are changed, so it happen a bit at a time. I may very well end up making the entire PC my self.

That is where I am at.

Dental labs use versions of (or have available to them) superalloys of the Inconel flavor. You should be able to make the parts in plastic or wax and have them investment cast under controlled atmosphere. I don't think that would be frightfully expensive.

I interned at a turbine engine R&D facility some time ago now. They investment casted the turbine blades that were in the hot section. They had tiny passage ways in them too. They would take them from the mold, polish / buff them, Weigh them, X-ray them, and if they passed, slide them into place and spool them up.

The casting process was VERY controlled to say the least. I still know some guys there and probably could get a few done, but that would only be worth doing for a final design version. I expect to make about three different PC versions before calling a design good enough. That is unless I discover something that will prolong the process.

I work with Inconel, Titanium, and Stainless more than carbon steel. I have to look through my notes to remind myself what alloy I wanted to use, but I am pretty sure I don't have any in stock. I wanted some really high temp stuff so I can keep the wall thickness as thin as possible to gain as much PC volume as possible.

At this point I am just puttering around with designs and theory in between other projects. I have 8mm and 10mm Elements, (Barrel and Plunger), on the way from different manufactures. Once I am confident that I have that project under control, I will start cutting metal for the first gen PC.

My over all objective will require several aspects of the stock engine be altered to match or take advantage of the changes to other specific parts. I do not want to scale up one process without equally scaling up the other aspects that are involved in the overall process.

All this is really just an exercise of the mind, and a challenge to the skills. I am not going to change the world if this all works out as planed or not.

A good friend of mine was looking for an engineering project to work on with his son over the summer vacation, so I told him about my Prechamber research project and he thought it would be perfect.

They are going to do advanced versions of the tests I have done so far with mechanical gauges. They will design a 24 bit, electronic data collection system with a very high sample rate to collect more accurate data.

Then they will take my simulated engine and collect some base line data which will include additional pressure and volume from, (simulated), combustion and boost. I have an idea how I was going to simulate the additional volume of combustion and boost, but I have not done it yet.

They seem to really be into the project, and I am very interested to see what they do with it.

I can say that my mechanical tests showed that the stock 616 Prechambers choke very quickly early in the RPM range, as in they do not fill very well above a certain piston velocity.

Bravo!!

I have been revisiting material options for the Gen II-Stage I prechambers.

I had originally considered 6AL-4V Titanium for its low thermal transference, corrosion resistance, and strength, but it's performance at high temperatures, (+500*F), is less than impressive.

As a result, 625 Inconel went to the top of the list, but it is expensive, somewhat difficult to work with, and is really overkill for a prechamber application.

After thinking about the prechamber being installed in the head, the Titanium’s lack of eagerness to absorb heat its self, the use of a thermo reflective coating on the inside of the chamber & burn tube and out side of the tip, and the fact that the high temperature compression and burn make up 50% of the operational time, leaving the other 50% for cooling, (dissipation), I am reconsidering the 6AL-4V. Another bonus for its use is I have a fair amount of it.

Additionally I have been leaning toward EDMing a helix in the burn tube and not using a ball as MB uses.

With the swirl having a vertical axis, I do not think there would be an advantage in angling the injector as there is when the swirl has a horizontal axis.

I am thinking that with high flow slots instead of the small holes, the helix will spin the higher flow volume fast enough to spread out to the outer edges of the chamber for good scavenging. This is where angling the injector may be advantages as it "might" create a compound swirl.

Testing the flow capability of the prechamber is easy, but I am trying to figure out how to design a test to see the quality of the swirl. It happens so fast it is going to be tuff.

Any thoughts on how to create a test to see the swirl quality in the lab?

Uhm.... no. The prechamber is actually EXTREMELY hot. Try 900*C (1650*F). See the figure I've attached for the temperature profile of the entire prechamber is. Even if you do a ceramic coating, etc, etc. I Don't think you're going to drop the temperature of the metal by 1000*F. FWIW, Mercedes uses Nimonic 80A for the prechamber.

That information and figure come from an SAE paper published by Mercedes titled "The Turbocharged Five-Cylinder Diesel Engine for the Mercedes-Benz 300SD". It's an excellent read, and I highly recommend it!

Attached Thumbnails

 

__________________
John Robbins
'05 E320 CDI - 240k
'87 300TD - 318k

I have seen the report, mostly scanned over it. There is a lot of good information in the report.

Thanks for the material identification that MB uses.

If I loose the tip I should be good, otherwise if I end up using some kind of tip, it would appear that inconel would be the only material that I should consider.

Edit: "After looking at the specks for the 80A, I am comfortable with reworking the stock prechamber. I was concerned about the strength of the stock material, but it looks to be a very good material so I may do a hybrid prechamber for a proof of concept to show which way is better. "

One of the major reasons the temperatures at the tip are so high because of the restriction caused by the small holes. The friction of the gases account for a good portion of the tip temperature.

For one of my flow tests, I used a 2,500 psi tank of compressed air as a pressure source that represented the cylinder side of the chamber. I put a flow meter in place of the injector and a pressure gauge that measured the pressure on the cylinder side of the chamber.

I slowly increased the pressure on the cylinder side (500PSI MAX) and observed the flow on the injector side. I had to keep the tests short as the control valve would get hot, but it was also very evident that the prechamber was getting way hot as well. So much so that I put a thermocouple on the tip to get some measurements.


Some project background;

The Gen I-Stage I prechambers were stock chambers with the only the holes enlarged.

Gen I-Stage I flow tests were conducted and as expected the flow increased at a lower pressure than the stock configuration. Also the tip temperature went down about 25% for the same flow as was seen with the stock prechamber.

The Gen I-Stage II prechambers are stock prechambers with the tip cut off. I sold my test car so I have not run the Stage IIs yet. I am considering putting them in my driver if I get the time to do so. I am confident that the engine will be ok at idle & low power, but at full power I am concerned the boundary layer may not isolate the piston from the concentrated heat that will be jetting out of the Prechamber.

My Gen II-Stage I prechambers will be as close to an open end, (Gen I-Stage II), but with a deflector on the end to help shield the piston from concentrated heating.

From the testing that I and a friend have done, I can see that the elevated piston temperatures that are shown the report are partially a result of the increased temperature as a result of increased volume in the cylinder, (from Boost), that could not get into the prechamber because of the restrictive orifices. As a result the additional volume compressed on the cylinder side of the prechamber the temperature of the compressed cylinder gasses will skyrocket, heating the piston, prechamber, and head, while robbing power from pumping losses. This increased temperature will be further increase the prechamber tip temperature in addition to the friction heating during the transfer of gasses between the cylinder and prechamber and vise versa.

My thinking is loose the restriction and gain the lost power from pumping, reduce the friction heating during the compression transfer, not create excessive compression heating of the piston, reduce the stresses on the engine from excessive and non-productive cylinder pressures, allow the productive combustion gasses to get to the cylinder in a grater volume over a shorter period of time as to be used to produce power.

As to the reasons why the small holes are there, I understand. They have been discussed in this thread. (I say this not to be snotty, but to avoid rehashing preciously discussed information). I am not, and will not, be running a stock engine, I am not concerned about emissions, noise, or fuel economy for that mater.

These prechambers are part of a total system as I am also altering the injection pulse width, (duration) via custom 10mm elements)), and quantity fuel delivered along with the Injector pop pressure, injection timing, and amount of boost.

I am doing everything with an end goal in mind, so I want to test and confirm every idea, instead of just putting a bunch of butchered parts together and wondering why it doesn’t work as I want it to.

For the moment, just a brief remark on the alloy. Various super alloys are commonly used in the dental industry for removable partial denture frameworks. Some are also used as frameworks for porcelain fused to metal.

The temperatures mentioned are the normal ceramic firing temps for these sorts of items so that should not be an issue. There should be any number of labs around that can cast high purity examples of these alloys under controlled atmospheres, etc. for a relatively reasonable cost. The expense would be making accurate plastic patterns that would be burned out of an investment mold. I believe there are digitized methods of building up plastic resin now so perhaps that is solved already.

Try contacting these people for Ticonium alloys or leads to labs using them. Several of my instructors and co-workers were trained in the military and used Ticonium products and so we got to hear all about them. The Ni-Cr alloys are generally easier to machine as well.

http://www.cmpindustries.com/index.php?s=2

[Also check out Nobilium and Niranium on the same website.]

Xrays? I dont think anything else will give high enough shutter speed.

__________________
http://superturbodiesel.com/images/sig.04.10.jpg
1995 E420 Schwarz
1995 E300 Weiss
#1987 300D Sturmmachine
#1991 300D Nearly Perfect
#1994 E320 Cabriolet
#1995 E320 Touring
#1985 300D Sedan
OBK #42

A friend of mine was a reactor operator at U of M when they had their nuclear reactor running. One of the types of radiation was used to see fluids leaking past a seal, deep inside a transmission during operation, for Ford Motor Company. Pretty cool stuff, but the reactor was shut down several years ago.

With the 80A material identified, I am leaning more toward modifying the sock lower half and making a new upper half with angled threads and an increased volume. I also want to look at a 60X prechambre as it may be easer and faster to adapt and modify one of those for my 616. Mine is a 78, so it has the early prechambers with the flat end, so I am going to make a tool to machine a pocket in the piston, (through the prechamber hole), for the "up rated" prechambers.

With a higher flow capability, the swirl in the stock prechamber would have to be increased I would think, so in the interest of saving time and maintaining as much known technology as possible, I think would be advantageous of me to mod a 60X prechamber.

Now for a prechamber for the 617a, a new prechamber would need to be made as the burn tube diameters are different, (larger), than the 60X prechambers.

In that instance, Kevin's suggestion of the dental lab for casting small numbers of parts looks like a good path to take. However the quality of any casting would be a concern to me.

It would be great if I can find some used 60X prechambers with the angled injectors to use a ginipigs. I also would like to have an injector for the angled prechambers.

I don't know what you hope to achieve with a swirl around the vertical axis. The whole purpose of swirling air in a prechamber is to circulate as much fresh air past the burn zone as possible. After the injected stream hits the impingement ball it will start by burning near the the ball force the burn outward as the O2 in the center is used up. Centrifugal force will also fling fuel against the wall of the prechamber where it will be further deprived of O2 and the heat will be absorbed by the wall. You will have no mechanism to swirl the air below the ball up where the fuel is burning and not improve air circulation beyond the stock setup.


Ideally you don't want to burn fuel near the walls of the prechamber because they take the heat away from the combustion gasses. The angled injector prechamber gets some of its gains because a lot of the hot burning fuel air mixture gets forced down the flame tube and thus less heat from combustion gets back to the prechamber walls.

__________________
green 85 300SD 200K miles "Das Schlepper Frog" With a OM603 TBO360 turbo ( To be intercooled someday )( Kalifornistani emissons )
white 79 300SD 200K'ish miles "Farfegnugen" (RIP - cracked crank)
desert storm primer 63 T-bird "The Undead" (long term hibernation)

http://ecomodder.com/forum/fe-graphs/sig692a.png