EngMod2T Q&A

[dmcca]

So ive got some questions about Engmod2T... hopefully there are other users on KB that can help, and maybe we can create a bit of a database of info to share. The program seems very powerful but i sometimes wonder if im feeding in accurate data or if im getting things a little wrong... without a dyno at hand its very hard to be certain.


When modelling a reed cage that has an internal divider or stuffer before the reed petals, do you subtract the area of the divider from the area of entrance to the reed cage when calculating the 'diameter of the rubber block' (entered in the Inlet Port window)?

The photo below shows what im talking about...

The rubber intake boot is on the right, the reed cage is on the left.
The rectangular area at the exit of the rubber boot is calculated and entered as the last diameter of the intake pipe (basically modelled as the cylinder side of the carb). In this case the area is 1190mm^2 which equates to an equivalent diameter of 38.9mm. This is entered in the Inlet Type window.

When you then go to the Inlet Port window to model the reed cage itself it asks for the 'diameter of rubber block'... this is essentially the area at the entrance to the reed cage but do i just enter the same value of 38.9mm diameter or remove the equivalent area (in direction of flow) of the divider from this figure??... in this case the divider is 8.5mm x 39.7mm = 337.45mm^2, so the total area of the reed entrance becomes 1190-337.45 = 852.55mm^2 or equivalent diameter of 32.5mm

The graph below shows a KTM200 modelled with the above reed cage, black line is a "diameter of rubber block" value of 32.5mm (divider removed from total area), red line is a "diameter of rubber block" value of 38.9mm (divider ignored). Obviously taking the divider into account looks better, but which one is actually correct???



To complicate things further, in this case the divider protrudes from the reed cage by about 10mm, so when its bolted to the rubber intake it technically changes the final area of the intake pipe just before it ends... should i actually be removing the area of the divider from the last stage of the cylinder side carb pipe instead of from the entrance to the reed cage???

[wobbly]

The inlet port and reed is very tricky to model correctly.
Look at the RZ400 inlet for my F3 bike.
The 39 dia is the actual area at the reed block face ( less the divider ), the 41.5 dia is the effective reed block flow area.
The 35mm is 2/3 of the reed block length, this models accurately the stuffer.
At the other end is the carb trumpet with a 6mm end correction length, this 6mm at 69mm dia is the flow area curtain at the carb entrance.
Using this setup you get the correct effective length of the inlet and the waves bouncing up and down the tuned lenfth are timed correctly.
In reality my "rubber boot" is bigger in diameter at the reed face, but using the effective diameter less the divider sets the actual area as seen by the flow.

This info was gained off an instrumented engine used for a PhD project at AK Uni, and initially I left the reed length off altogether, and added 1/2 the carbs trumpet dia at the other end.
This gives approx the same length overall.
But the method above is more accurate for all engine and reed sizes.

[teriks]

Good info, to bad I'm not dealing with reeds (yet anyway). On the other hand, modelling the atmospheric side is hard enough to deal with.
Do I understand you correctly if I say that D=69 is what Neels calls "Starting diameter" in the help files?

Now to something that's been bothering me for a while, how to model Prof. Blair's "Best Bell"?
(More Blair candy at http://www.profblairandassociates.com/RET_Articles.html, mostly four-stroke thou.)
Judging by the article, I would believe that the correction length should be kept well below the 0.5 times the starting diameter suggested in the help files.
-It looks like you have come to the same conclusion on your RZ400 model, although I don't know the geometry of your carb.

Apart from a bunch of simulations I have built and ran such an intake, but without a dyno it's all but impossible to evaluate the small improvements suggested by simulation.

Need some sleep, but Ill try to produce something describing my modelling tomorrow.

[wobbly]

Yep, the 69 is the "starting dia" of the bellmouth.
The trick is to use the correction length that gives you the effective curtain area equal to that start dia area.
The reed end is alot more complex, as they are open/closed/partially open all the time, so present a variable length during one cycle.
But using the lengths described gives an accurate representation on screen,that correlates with the instrumented engine testing.

[teriks]

So, as promised, this is how I approached the "best bell".
Black lines=actual geometry.
Red lines=simulation model.

Seems I have put way too much detail in the model, both compared to wobbly's example and to the examples in the help files.

Now, how's "effective curtain area" defined?
If you said just curtain area, I would have guessed 22*PI*4 in my example above, but now I'm lost.

Can I use the fact that I was blond as a kid as an excuse for asking stupid questions?

[wobbly]

I dream about blonds so no you cant, it would ruin everything.
In your model the sim would work well with a taper from 12 out to the 22 dia bell entrance face dia.
The end correction is defined as the carb dia squared divided by 4 times the bell entrance dia.

[teriks]

I think I'd rather don't mess with those dreams

Now that's a definition of correction length that makes some sense even to me, thanks wobbly, this is gold!
Made some sort of illustration as a reference for my self, might as well share it.

[teriks]

Double post...

[dmcca]

The 39 dia is the actual area at the reed block face ( less the divider ), the 41.5 dia is the effective reed block flow area.

This info was gained off an instrumented engine used for a PhD project at AK Uni, and initially I left the reed length off altogether, and added 1/2 the carbs trumpet dia at the other end.
This gives approx the same length overall.
But the method above is more accurate for all engine and reed sizes.

The end correction is defined as the carb dia squared divided by 4 times the bell entrance dia.

Thanks for the great info wobbly!!!

Is the effective reed block flow area you used just the area calculated by the program in the Inlet Port window? Or are you calculating a different figure for this?

[wobbly]

Yes, the reed program calculates the effective round port diameter, and I use this as the area at the end ( beginning) of the inlet length.

[wobbly]

Here is the single most important trick to be learnt about intake tuning.
This is RZ400 for F3.
The intake tuned length is optimised for 10,000 rpm, this is shown by the red line, case pressure, dropping below atmospheric, at the same
time as the intake pressure ratio yellow line,goes positive at the reed face.
This puts negative on the outside of the reed and positive on the inlet side, opening the reeds very quickly and early.
Shorten the intake and the yellow crossing point moves to the right,giving the coincident crossing point at higher rpm.
This gives the lie to the fact that using intake rubber and carb lengths off a 125 are correct for a 200 to 250cc cylinder as is done by most manufactures and is accepted by tuners as OK.

[dinamik2t]

Hi, I'm new here! (had a subscription for while, but never posted)

I 've been using EngMod2t for about a year now. I have previous experience from mota.. that's day vs night! I haven't had the chance to dyno test any results yet though.

I would like to share EngMod files, even though I don't have a lot of them. Mostly a couple of pipes, some carbs and a few moped tuned engine files. All 125cc class.

I also have a question. What value do you use for 'combustion efficiency'?? In the manual, mr Neels says that 0.87 should be right for modern engines and up to 0.92 for GP engines.
With 0.87 I get very good results, but I am not sure whether that value is a little optimistic.

Also, the engines we tune around here are mostly air-cooled. I see you modify air-cooled engines too. What values of cylinder temperature do you consider "safe" ?? I try not to exceed 2700 in sim results, but I 'm not sure about the real-life correspondance of these values.

[teriks]

Ok, so I have tested three different bell models.

RED: According to my first picture posted in this thread, end correction 4mm.
BLACK: #1, but with end correction 1.6mm.
GREEN: Wobbly version, my second picture posted in this thread.


Results:

The seemingly small change in end correction of 2.4mm gives a larger change in the result than changing the detailed model into a simple cone.
-If the 1.6correction is correct, I have work to do...

Perhaps I should bring this over to the ese-thread, lots of engmod-discussion there now.

[ief]

Such a good idea this thread and so little follow up...

I find it pretty hard as well but it's realy a great (and fun) tool, to bad about the only one I know of that knows what he is doing with it is Wob and well, Neels himself

I am playing with tuning the inlet myself now and found the opposite of what Wob was saying, thus, making it longer moves the wave to the right And then there is that making the inlet longer makes the wave less strong...

Thing I thought about was, there is only one sort of slide carb but was led to believe a flatslide performs differently from a round slide, that's not an issue?

Long story short, it will take quit some time for me to grasp this al a bit but that's part of the goal and fun.

On to modeling a carb correctly...

cheers!

[ief]

Or an airbox...

leeds to a question I came across earlier.

When looking at these two graphs I can't find the logic.

Ab6 = without airbox, makes less power (say 1 hp)
Ab7 = with airbox

Ab6 shows more reed disp. be it later, it closes later too so in my mind it should flow more, thus making more power. All makes sence, it's better timed and allthough the puls action is slightly less the proper timing compensates witch shows in the reed disp. So far so good but...

Ab7 makes more power?

So ithought I'd see what the flow at the inlet does and this baffles me even more, Ab7 seems to flow far less then Ab6 but still makes more power and reed disp and crank P ain't that different?

I must be missing something...

attached P traces M traces and combined.