ESE's works engine tuner

[teriks]

Attachment 251116

STA numbers for the graph, looks like plenty of Blow Down Time Area, I think it must be something to do with the exhaust tract/pipe.

I'm anything but an experienced tuner myself, but I believe you will always see more or less of that pressure spike around max-power-rpm on a useful engine. How much is too much is another question.

Re. the pipe, wobblys comment

The port timing and the pipe length are not working together, you have no superposition on the far left.

applies to this one as well. Compare the pressure in the exhaust port, "Pexport", at EPO with wobbly's pic here

No ... just learning, thanks for the tip.

So am I, after plenty of time behind the screen. It's kind of humbling...
Let's see what the more experienced have to say.

[wobbly]

Dont get hung up on Bl - STA, thinking its good to get rid of the transfer blip..
Any well developed race engine is ultimately Blowdown restricted, and you will always see the Tr Port pressure ratio rise when they open.
That effect is what makes port stagger work.
The first port to open, initially has backflow, it must, as there is more pressure above than below.
Then the rest of the ports open and eventually begin to flow into the cylinder due to the depression created by the pipe diffuser.
Thus the port that opens first, flows last.
Counter intuitive, but reality, proven on motored, instrumented engines in the lab and reported in a raft of SAE papers - and now shown in the sim.
Getting a balance between the blowdown needed to achieve the power you want, and the transfer area available is the tuning trick most valuable.

Superposition at ExPort opening is loosely referred to as pipe/port resonance, and is best achieved over the widest range with low Ex durations down at 192*.
Unfortunately this is countered by needing alot more blowdown than these numbers allow, and around 198 is needed to achieve the best bmep numbers.
In this scenario we try to achieve resonance at peak and beyond, to give plenty of overev power.

Next issue that is my favorite hobby horse for today is Ex duct exit area.
You will find a heap of free power, when using a T port or a Tripple port configuration, by limiting the duct exit area to around 75% of the Total Ex Port Effective.
Then make the header start at the area = to the Total Effective.
Join these with an oval to round adapter in the spigot or flange, where the width = the header dia all the way thru, to enhance the flow from the blowdown area,and
the height at the flange face forms an oval to give the correct 75% area.

The Temp Av in the pipe should be around 500 as in any fast engine you would measure around 600 in the header.
The TuMax should be around 1000,any more and you get deto, alot less and it means the temp/pressure rise in the chamber isnt high enough, and you are restricting the "push" on the piston.
But can also mean that you are using retarded timing to get heat into the pipe, not create pressure on the piston.

Lots more stuff, so little time.

[bucketracer]

Great info, thanks Wobbly, I am learning more all the time.

[dinamik2t]

As an addition to what Wobbly said about the exhaust duct-flange diameter, there is this quote from pitlane.biz:

One wants the pressure wave coming back into the cilinder but NOT the burned gases that are hot and can cause detonation!

and the photos of RSA's duct/flange

[F5 Dave]

Indeed. Makes the staggered port make more sense now.

Thank you Frits, lovely cases, esp the water channel presumably as a neat to barrel to barrel water transfer & taking heat out of the cases as well. Never was sure how much cooling the cases did. Seemed popular ages back & then just sort of went quiet (But I'm obviously not poking around high end GP bikes).

[kel]

Below are some pictures of the Aprilia RSA125; the red arrows indicate the ignition sensor.

Wow that RSA has some different bits attached? Frits is that a temp sensor in the exhaust header flange and if so was it connected to the engine management or just the data acquisition system?

[Frits Overmars]

Wow that RSA has some different bits attached? Frits is that a temp sensor in the exhaust header flange and if so was it connected to the engine management or just the data acquisition system?

Yes, it was a kappa-type thermocouple and it was just for data acquisition.
In fact its position is much too close to the cylinder; there should be room for at least one cylinder volume of 125 cc in the header between cylinder and sensor to avoid fresh charge hitting the sensor. But mounting the sensor further down the pipe proved impractical because there it was quickly destroyed by vibrations.

[dinamik2t]

Here are some results regarding inlet length tuning, to add to what Wobbly described in previous posts.
I am speaking of a 118cc with max hp at 11.8k rpm.

Firstly, an inlet pipe of total length of 136mm. (2/3 of reed cage included) Fine tuned at the "3rd harmonic", with inlet pressure rising and case pressure falling. Would be a charm, unless I would have to skip the whole intake rubber.


Then, same setup with the shorter intake that can fit into my chassis. Total L of 186mm. Tuned at nowhere (?) Same peak power, but of this more to the end.


I tried to tune it half-period earlier and here there are the results for inlet pressure droping while case pressure does too. At 161mm. Slightly decreased peak-power.


Finally, if I wanted to achieve the inlet-fall/case-drop situation with a realistic intake, I had to tune at the "2nd harmonic", with a length of 221mm. A small decrease of peak power here too. [Notice here that reed pressure rises significantly more than in the other two setups. I guess this might have an impact on reed life. - That is not correct, REEDDISP is Reed Petal Tip Lift, as Wobbly corrected]



I wanted to do an overlay image but there are too many graphs and it would be more confusing than helpful.
Anyway, those were the results at peak-power rpm only. And differences there where not very serious.
If we look at the whole curves though, things are very different. (the 3rd result is missing, I didn't run it)



-------------


Also have a question for you Wobbly, whenever you find the time to answer.

I though of simulating the Flange to Pipe Header transition, so I added a small part in front of the exhaust.
Starting diameter is the duct exit diameter and end diameter the header diameter. The length of the transition, I used to add it to ex duct length in Exhaust data, as the flange base thickness.
Is this thought correct or not?



I experimented a little with lengths and the setup in the image worked best for me.
Generally, comparing transition with no transition, there was a loss in overrev power and a small gain just before peak power.

[wobbly]

Firstly you are misguided re the reed trace - this represents reed tip lift, there is no "pressure ratio" being shown.
The reeds will flat line at the max lift if you use a stop plate.
The length of about 135 for a 12,000 rpm engine has an effect over a much wider range, than the longer intakes, as the pulse frequency is greater, and the crossing point doesnt
move as much with a change in rpm.
That length is what was achieved in the dyno graph I posted back a bit for the TM125MX - just taking 15mm off the rubber gave 2 Hp in 50, and worked over a wide range.
It used a SPJ short carb from a HRC RS125, and a very short rubber off a earlyer model with the same bolt pattern.

Re the Ex duct nozzle.
The area at the cylinder duct exit should be around 75% of total Ex effective, simply change this in the Ex Port page.
Then you should have at least 30mm or so of "flange" to increase the dia up to the header size ( this would be an oval to round CNC transition).
Your small taper is way too steep, and kills energy in both directions.
This flange is part of the pipe, as you have done.
When done right this is the result - no nozzle = a 45mm duct exit straight into the header.
The curves above had a 39mm dia duct exit, with a 30mm long flange going up to 45 dia at the header.

[k14]

As an addition to what Wobbly said about the exhaust duct-flange diameter, there is this quote from pitlane.biz:

and the photos of RSA's duct/flange

Makes me remember when I was racing 125GP and plenty of guys got the devcon out on the header flange to correct Hondas "mistake"

[dmcca]

Re the Ex duct nozzle.
The area at the cylinder duct exit should be around 75% of total Ex effective, simply change this in the Ex Port page.
Then you should have at least 30mm or so of "flange" to increase the dia up to the header size ( this would be an oval to round CNC transition).
Your small taper is way too steep, and kills energy in both directions.
This flange is part of the pipe, as you have done.
When done right this is the result - no nozzle = a 45mm duct exit straight into the header.
The curves above had a 39mm dia duct exit, with a 30mm long flange going up to 45 dia at the header.

Sorry if this is a stupid question... but do you want to have a step at the top and bottom where the flange/duct meet?? I imagined the CNC oval/round transition to be smooth but it looks like its not the case in the RSA photos above???

This seems like a key bit of tuning advice so i want to be 110% sure i get it. Ive mostly worked with MX motors and most have this 'step' top or bottom or both

[husaberg]

Question for the panel

In the Bell book he mentioned modifying the two crankcase ports in a CR125 cylinder (the old cylinder reed model which has two small ports with link the reed cavity with the crankcase)
He closed them with filler and then shortened the piston skirt to achieve 230 degrees inlet duration.

I did the same mod on my old H100 which has similar ports and a windowless piston but never got around to testing it.
I always wondered how much difference it would actually make and why.

Has anyone ever tried it?
He said he achieved an extra HP or so with the mod.

[dinamik2t]

Woobly, merci. I read "reeddisp" -displacement probably- but then again, all Y-axis was pressure. Thanks for making that clear.
-I edited the previous post so as not to mislead anyone.

I have a 37 efficient diameter at ExP and duct exit was 36. Before the above sims, I have used every duct exit size from 28(~75% of 37) till 36 - 2mm stepping. Best results in my opinion were at 32mm. I have the graph for the 28~34 testing testing. At 28 there was a huge overrev gain, but power decreased everywere else.



Actually, now that I look back at your post about RZ400 numbers, considering that header dia. is the same as ExP effective dia., then 39 is ~85% of 45.
About 85% is the 32 of the 37, which gave better results in my scenario also. Perhaps you are too busy these days

Sorry if this is a stupid question... but do you want to have a step at the top and bottom where the flange/duct meet?? I imagined the CNC oval/round transition to be smooth but it looks like its not the case in the RSA photos above???

I remember Jan (Thiel) or Frits (Overmars) -- saying that duct volume should be as small as possible. Plus the quote of Jan's words earlier.
I don't think this gives you the answer you want, but anyway. We should try and think how a wave behaves to understand further of this configuration.

[dmcca]

We should try and think how a wave behaves to understand further of this configuration.

Thats what ive been wondering... taking the quote earlier from Jan Thiel into account (about wanting the return pressure wave but not the returning gasses) I would guess that the step hinders the reverse flow of the burnt gasses , possibly by bouncing back a small wave back down the pipe as the reverse wave returns to the exhaust duct??

*** on second thoughts the 75% oval to 100% round transition will create a reverse wave just fine by itself (even without the step) as the initial wave returns though it.

[teriks]

I cant really come to terms with the step either, but it would hinder _flow_ from the pipe to the cylinder somewhat.

The effect of duct volume, or rather duct cross sectional area, I think I have some kind of grasp on.
-Decreasing the duct cross section area would increase the length of the fresh gas column for a given volume.
This could increase power when the return pulse otherwise arrive too early.

I'm sure Frits or wobbly will correct me if I'm wrong