Fuel stand-off - intake reversion - what are the causes? 2

[v114]

A while back, I posted some questions related to a v-twin I am working on.

Here is another puzzling one. The engine is a v-twin, dual overhead cam, 4 valve configuration - designed from a clean sheet of paper - not really a copy of anything.

The motor has severe intake reversion that comes on at around 4-5k rpm. At the same time, the engine will not make peak power past low 5k rpm.
Things tried have been:

1. Cam timing and duration changes - we have been all over the map with possible changes and combinations and the engine does not really respond positively or negatively to cam changes.
2. Intake manifold configs: dual IR runners, one carb to divide to either cylinder, 48mm and 56mm throat carbs. All had reversion.

Is it safe to say that the engine will not be able to realize its power potential unless the reversion goes away? Also, what other factors could contribute to reversion besides cams and intake systems? Really look forward to some experienced opinions on this topic.

Thanks

 

[patprimmer]

Intake runner length.

Intake runner cross sectional area.

Intake plenum volume.

Design and shape of bell mouth to runner.

Throttle plate position and size.

Intake valve closing point.

Valve springs and cam closing ramps and their effect on valve bounce.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[BrianPetersen]

What RPM is the engine designed for? What's the volume of each cylinder? What type of carb are you using - automotive fixed-venturi, constant-velocity with the slide actuated by venturi vacuum (standard motorcycle carbs), or flat-slide-type carbs in which the slide and needle also act as the throttle?

My experience, which is motorcycle engines with a separate carb / throttle body per cylinder, is that you will *always* have some degree of flow reversion, particularly at low revs. If the cam timing is long enough for the engine to make decent power at high revs then it will be long enough so that at lower revs it fills the cylinder first and then pushes some of it back out. If you run the engine without the airbox you can often see a mist of fuel being pushed back out.

If your problem develops at high revs and full load, it sounds like there is some sort of resonance developing between the intake runner and the cylinder (acting like a Helmholtz resonator). I can only see that situation happening if your intake ports or intake runners are WAY WAY too big and/or too short. You need to keep the velocity up so that the column of air stuffs the cylinder full but doesn't let any of it back out when the engine is in its powerband. I've never seen a production-design engine have that problem, but if you've gone out of bounds, maybe there's a first time.

The bike engines are just designed to deal with whatever reversion happens at lower revs. Constant-vacuum carbs have the slides close down at lower revs, which partially reflects the reversion wave; the carb is jetted to deal with whatever reversion is left. Modern fuel-injection systems use a drive-by-wire throttle that stays mostly closed at low revs, even when the rider asks for full throttle. Flat-slide carbs don't deal with the reversion; they have terrible bottom-end response ...

 

[v114]

Some of the comments included so far, we have already experimented with.
I'll explain.

The design is a 4.25 bore, 4.00 stroke, 6.8" rod, 45 deg twin. We would expect to see the best power in the 6500 to 7000 rpm range. It seems that the reversion is not a problem "visually" at under 3500 rpm. Once over 4000 rpm at full throttle it comes in and get's worse and worse as the rpm climbs and never goes away.
On the dyno, the mist of fuel/air coming out the carb mouth is around a 1.5 ft column.

In terms of determining if the intake runner cross section is too large, we have gone from as low as 1.8 id to as large as 2.25 dia just after the carb body. In fact, the reversion seemed worst on the smaller runner.

Since the motor fits a Harley style frame, the intake runner length is fairly short in order to be packaged. But, for tests, we have tried adding a spacer up to 3" long behind the carb with no better results.

The carbs tried have been 48mm mikuni flatslide and S&SD butterfly style with 2.25" throat. Both had reversion at high rpm.

The intake cams have been tried as low as 226 degrees @.050 duration and 97 degree centerline (early closing point). This was not better than larger cams with later closing - makes no sense so far.

Brian, it sounds like your experience is that the reversion goes away once the revs come up. That is opposite of what we see.

Is it possible that the bore and stroke will not work well with the head flows we are working with on the 4-valve head design? We have very good low lift flow numbers as expected with the 4 valve design. Does the motor need to be more over square?

Pat, I remember you replied to my "optimum port velocity" questions a few years ago. You mentioned that the intake cam should guide the valves to meet the piston and follow it down the bore and close as soon as can be done to achieve best power - can you explain further?

I remember you also stating that the intake runner volume should be slightly larger than a comparable 2 valve. Could you give recommendations for sizing based on this 114 cuin engine?



What keeps the motor from carrying out on top end?

Thanks for the help

 

[automotivebreath]

One thing to look for is incomplete combustion. If the flame
speed is too slow it will continue to burn during the camshaft
overlap period causing excess pressure in the intake tract.

 

[v114]

Automotivebreath,

This isn't something I've even considered. How do you go about finding if this is happening? What test can be done?

 

[BrianPetersen]

Okay, that's useful info to compare to my experience.

What's happening on the exhaust side, and how much cam overlap do you have? If you are getting a pressure wave coming back the exhaust system during the valve overlap period, it's going to spit flow backwards even before the intake stroke starts. If the exhaust system is too restrictive to let the flow out, there could be pressure left over at the end of the exhaust stroke. This gets *worse* if the valves have high flow at low lift.

Back-of-notepad calculations suggest that your exhaust system should have primary header pipes somewhere around 40" long and 2.5" diameter. On a 4-banger you would want to join with the pipe from another cylinder at that point and carry on with about the same diameter for some distance after that.

The short intake runners are unfavorable, but with that engine configuration I agree that there's not much that can be done.

Are you drawing air from a good-sized airbox? Most bike applications just contain the reversion flow and have the engine breathe it in again.

 

[BrianPetersen]

One other thing with the incomplete combustion situation. The engines that I deal with range from 56mm to 76mm bore or thereabouts; 76mm is about 3", but the rev range is about double yours. It should be *possible* to get a good burn if the combustion chamber is right. If it's a 4-valve head, one would assume you've put the sparkplug dead centre, and have a squish-band around the perimeter of the chamber everywhere that there isn't a valve taking up the space, and hopefully you've used a low included valve angle and either flat-top or dished-centre pistons ... or not?

 

[automotivebreath]

Short of pulling the heads for inspection several indicators
can be used. Excessive EGT, intake inspection, spark plug
inspection and exhaust inspection. (dry/white deposits = good
combustion heat; wet/black deposits = insufficient combustion
heat)

Here’s a good ink to reading plugs:

http://www.empirenet.com/pkelley2/sparkplugreading.html

Another thing to consider is exhaust system efficiency;
properly designed it should pull on the intake tract at RPM.

 

[FoMoCoMoFo]

agree, once saw an engine behave just like that. turned out to be restrictive exhaust. hard to imagine that happening in this case though. you probably already know...it going way rich when reversion gets that severe.

 

[v114]

The air/fuel ratio is in the 12.5-13.0 area for best power. At this setting, the plugs look white/slightly tan and dry.

The head/chamber configuration is modern - plug in center, shallow valve angle and squish band around where we could get it. Spend lots of time on this.

The motor seems to make best power with only 26-28 degree timing.

Brian,

I have tried cam timing all over the place including large and small overlaps - it doesn't really seems to matter. This the confusing part. I believe that there could be something in the way of being able to see the results of cam changes.

The current header pipe is 1.9 id and around 38" long through flow thru mufflers. This is the largest header pipe we used to date. The outlet of the port is 1.800.

On the intake runner, would it make sense to try to build an automotive style plenum "box" after the carb to increase volume understanding that there will be a sudden drop in velocity as air passes into the box area. There really isn't room to make the runner longer, but it could be larger in diameter before the heads.
This could contain/diffuse the reversion, but doesn't it make more sense to try to cure it in the first place?

 

[thundair]

One thought is to see if you have a lazy port so...check velocity in both ports.
The other is (this may have been mentioned) that a reversion wave is timed to the RPM, if the positive wave hits the chamber during overlap it could spew out the intake....
Dont try to fix it with an air box as it will be miserable to tune with a sometimes wet reversion wave...

Cheers

I don't know anything but the people that do.

 

[thundair]

Is there an edit?...I meant a positive exhaust wave into the chamber and spewing out the intake..


Cheers


I don't know anything but the people that do.

 

[jimwolf]

I no nothing about v twins, but I am familiar with quad valve 4,6,&8 cyl. The fog ball is not that unusual for 1 throat/cyl carb setups, but if it is excessive, the only thing you said that may cause some issues at higher RPM is the tight lobe center (97deg) for the duration cam you mention. I am guessing that the cam is around 260 deg. actual which is about normal for a cam that is 226@.050". On engines I am familiar with, 112 LC is closer to typical for a well mannered engine and is a long way from 97deg LC. You didn't quantify how far you had spread the lobe centers in testing, but I would be more inclined to start testing in the 110 to 114 lc area if the fuel reversion is a problem. Watch your piston to valve clearance when moving the L.C.s around this much.

 

[patprimmer]

He still has not qualified intake runner length or whether it's one or two carbies.

The stand off should tend to suck into the engine as revs and airspeed increases.

I agree 112 lobe centres on a 4 valve head. They scavenge a lot better than 2 valves and to much overlap increases the chance of exhaust pulses bouncing back into the inlet, although you should be able to drive through these events as they should only occur over a narrow rpm range.

It really sounds to me like cam timing is way out, like it lost a lobe type out.

I guess it's to basic to ask, have you checked the split overlap is at about TDC.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[v114]

I really appreciate the feedback - might just make some sense of this after all.

1. Runner lengths - intake port - 5.5 inch average between the two valves. one runner is slightly longer than another.

intake manifold - 5 inch average - one cyl in further than another because of cyl offset.

total avg runner length - 10.5 inches.

2. It is a single carb, single throat 57mm, butterfly style. The carb feeds both cylinders through a common manifold which divides to each cylinder.

3. I have not checked the split overlap, but I can offer this information:

Recent cams and settings:

1. intake 226@.050 set at 97 deg lobe center
exhaust 234@.050 set at 112 lobe center

lobe center separation = 97+112/2 = 104.5

2. intake 234@.050 set at 100 lobe center
exhaust 226@.050 set at 106 lobe center

lobe center separation = 100+106 = 103

3. intake 234@.050 set at 100 lobe center
exhaust 234@.050 set at 112 lobe center

lobe center separation = 100+112/2 = 106

All of the above do not come very close to the 112 lcs described above. Could this be an issue? Do the above figures point toward a tendancy toward reversion at higher rpm's?

If yes, then would a setting the like the following be more in line

intake 226@.050 set at 108 lobe center
exhaust 234@.050 set at 116 lobe center

lobe center separation 112.

 

[patprimmer]

I guess it is twin cam so you can dial in each.

I would go back to basics and set it up at split overlap with 112 Deg between centres.

Fine tune from there.

Durations sound in the ball park to me.

Inlet tract is a lot to short.

Can you fabricate something up for dyno that makes both runners the same length, with a connecting plenum. Make the plenum quite large and put the carby in the middle, possibly facing away from the runner inlets so the airflow sweeps through a 120 deg or more curve. This might keep the reversion on the plenum and make a bit of room for longer stacks.

Are the valves definitely holding their seats without leaking. Is the head distorting under load and unseating the valves.

This is all grasping at straws, but your problem really does not add up.

Can you test with long, big dia open exhaust pipes.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[dicer]

I've seen this on F1 engines, and they seem to do okay with it.

 

[patprimmer]

Are you sure about your real in cylinder A:F ratios.

Maybe it has good mixture with stand off leaning it out, but as you draw the stand off in, it is going rich and miss firing.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[v114]

Pat,

Very much appreciate your suggestions.

The a/f ratios are being read by an oxygen sensor on the front headpipe.

The next step for me is to try the 112 centerline intake and exhaust.

On the plenum idea, I think a "box" can be made up to bolt onto the existing intake manifold flange. 90 degrees to that inlet, the carb can feed through into the box. This should point the carb throat toward the front of the bike into the airstream - a good thing.

Have any ideas on what volume the box should be in relation to the engine size? Something like around 50 cu in should be possible to package. Wouldn't there be any concern about the sudden drop in velocity of the air once it first hits the plenum behind the carb?