Cam duration for slow running four cycle model engine

[GreenTwin]

I would think that without a dyno test, one would never be able to distinguish one cam from another, other than perhaps with how smooth the idle was at slow speed.

I found the torque curve for my Yamaha SR500 (one cylinder, 500cc four stroke engine), and it is perfectly flat, which according to the engine designer was intentional. A flat torque curve makes the bike a joy to ride, and often eliminates a lot of shifting, since you don't have to stay on top of some narrow power/torque curve like some of the multi-cylinder bikes I have owned in the past.

My Honda CR500 (one cylinder, 500cc two-stroke) engine has a very light flywheel, and a very narrow power band.
It is what I call an "on-off" motor, where you are basically at either zero horsepower, or 50 hp.
The throttle is 1/4 turn.
I added a heavier flywheel, which had no effect at all.
The CR500 is set up to win races as it is furnished from the factory, and so it does not do well just poking along in enduro-style.
I basically put up with the CR engine in order to get the befits of the superb suspension system, which is 12" of travel front and back, and a non-linear single monoshock in the back. The CR is 230 lbs total weight, and so you have to hang on and be cautious with the throttle if you don't want to end up on your behind. The power-to-weigh ratio is phenomenal.

I have an electric scooter that I ride, since I had to give up running due to knee problems.
I would like to make a small 4-stroke motor for it, to eliminate the charging, and avoid the ultimate battery replacement that will be required one day.
I only need to go perhaps 12 mph max., but I need torque for hills, and I need something very quiet, basically as quite as the electric version I have, or close to that.

So I am reading this cam info with great interest, and taking copious notes.

.

 

[Brian Rupnow]

Thank you George. I was hoping you would chime in on this subject. As you can see in the cad overlay that I posted, the difference in a 120 degree cam and a 130 degree cam is very negligible. Oh granted, there would be some difference in the way the engine responded to a change like this, but I doubt very much that unless you were designing for extreme torque and power, there would be very little difference how it affected the engine.---Brian

 

[GreenTwin]

How does a dyno measure torque and horsepower?

I have seen the simple brake with gauge, which I guess measures torque?

.

 

[peterl95124]

View attachment 140100
Yours is the 'up' triangles while Lotus's default is the 'down' triangles. The 'super slow' 100 degree cam remains the crosses. Based on this, your cam eliminates the downsides of too much overlap seen on the stock cam, but still isn't a good match to this engine's intake and speed range when compared to the 100 degree cam.

many thanks. mine being an aircraft engine I'm not trying for low RPM power, but since I only have a wood prop on it that's been drilled out for the 1" diam hub, I'm sticking with a 6.5mm carb even though I have 7.5 and 9 on hand. once primed it always starts on the first flip, and I'd like it to be capable of 3000 / (scale factor) ~~> 15,000 RPM, but I'm not going past 4 or 5 K with the wood prop . So I'm very satisfied with this cam. if you have any additional thoughts or advice feel free to comment.

 

[peterl95124]

I have a diagram I got off the internet a long time ago showing how the Variable Valve Timing is adjusted in the Toyota Prius hybrid engine for various driving conditions, which only varies the intake timing, the exhaust timing is fixed. It was scanned at too low resolution and the text is on the verge of unreadable so I'd like to try to clean it up before posting it here, but the upshot is

intake delayed for: starting, stopping, idling, low temperature
intake advanced for: (3)medium load, (4)high load at low to medium speed,
intake in the middle for: (2)light load, (5)heavy load at high speed

the reasons given are bewildering...
(2)decrease overlap to eliminate blowback
(3)increase overlap to increase EGR to lower pumping loss
(4)advance intake close for volumetric efficiency
(5)retard intake close for volumetric efficiency

so its hard to draw any conclusions for model engines other than for easy starting you want no overlap, so I've stuck with that for all my model IC engines.

and its why I jumped to the conclusion that obsessing about cam angles for model engines was a waste of time, but Nerd1000's graphs show that you can tune a cam for low RPM power if that's your goal.

HTH, YMMV, etc...


PS, I've tried uploading the paper but keep getting "file too large" errors...

 

[peterl95124]

I found a camera that takes photos at less resolution than my (new) iPhone and doesn't get errors when I upload, so here is the Toyota Prius' Variable Valve Timing page I was referring to

if its not legible I can add comments

 

[Nerd1000]

The actual physics behind valve timing is that the gases flowing into or out of the cylinder have momentum, it takes some time for them to stop and start flowing. We can exploit this to improve volumetric efficiency in several ways: firstly, by closing the inlet valve after bottom dead center, the momentum of the air keeps it flowing into the cylinder even as the piston starts to rise, allowing us to pack more air into the cylinder than we would otherwise. Secondly, the exhaust gases have enough momentum that they don't stop flowing out of the cylinder immediately even after the piston reaches top dead center on the exhaust stroke, by opening the inlet valve before TDC and closing the exhaust after TDC we can set up a flow of gases that tends to blow residual exhaust out of the combustion chamber during this 'overlap' period. Obviously the faster the engine runs the faster the gases have to flow in and out, which naturally strengthens these effects by giving the gases more momentum. Cam timing is basically a matter of setting up the valve events to maximise the benefits of all of the above at a given engine speed. What's less obvious at first is that our model engines actually run at very low speeds compared to most production engines- 6000 rpm on a 25mm stroke is a lot slower (in terms of mean piston speed) than 6000 rpm on a 100mm stroke, so we must adjust our timing accordingly.

If you leave the inlet open too long the air will have time to stop and begin to flow back into the intake, while if you close it too soon you miss out on the inertial cylinder filling effect (either way, the engine loses volumetric efficiency and thus torque). This is the strongest cam tuning factor on most engines, which is why many older car engines use a cam phaser only on the intake side. Leaving the exhaust open for too long after TDC leaves time for the exhaust gases to stop and begin flowing back into the cylinder during the intake stroke, which has pros and cons (diluting the charge with exhaust compromises power, but it cools the combustion process reducing NOx emissions. It may also allow you to open the throttle more for the desired power level, which reduces pumping losses). For maximum performance it is also desirable to open the exhaust soon enough for most of the exhaust pressure to 'blow down' before the piston reaches BDC, obviously the faster the engine revs the less time is available for this so the exhaust should open sooner in the cycle.


The Prius is an unusual case because it is intended to mimic an Atkinson cycle engine by using an inlet cam that (by normal standards) has excessive duration, under most conditions the engine deliberately reduces its volumetric efficiency by pushing some charge back out the intake during the compression stroke. This allows good economy at light load (and an unusually high compression ratio for an engine that runs on 91 RON petrol- 13:1!) but is majorly bad for low RPM torque- of course this isn't a problem for a hybrid car, as the electric motor provides the majority of the torque required at low speed.

 

[Charles Lamont]

<...>As you can see in the cad overlay that I posted, the difference in a 120 degree cam and a 130 degree cam is very negligible.<...>.---Brian

Brian - I have already been at pains to explain, in posts 5 & 6, why your cad overlay appears to me to misunderstand the difference between a 120 and a 130 degree cam.

 

[petertha]

IMO it's silly to be worrying about duration or lobe separation angle. Therefore, a basic set of parameters for a slow engine might be as follows:

Inlet valve opening: 0 degrees (TDC)
Inlet valve closing: 20 degrees after BDC (200 degrees crankshaft rotation)
exhaust valve opening: 20 degrees before BDC
exhaust valve closing: 0 degrees (TDC).

Coincidentally I was working on a cam timing spreadsheet for another engine. I loaded your example to match the duration & timing relative to TDC & BDC. The only way I see it working is if the LSA (lobe separation angle) is 100-deg. LSA is represented on the chart by the line starting at 260 deg, ending at 460-deg which corresponds to EXH & INT cam centerlines respectively (their maximum lift points). To confirm, 460-260=200 deg CS angle. 200/2=100 deg between cam lobes.

I think it suggests LSA is equally important. For example if LSA is increased/decreased, the red/green EXH/INT bars diverge/converge accordingly. Even though the individual cam duration remains unchanged, the resultant timing angles relative to TDC/BDC will be different. For example, if one just assumed a right angle 90-deg LSA without knowing better, it would be physically impossible to achieve this particular timing given the stated cam durations. Does this math go around or am I missing something?

 

[Nerd1000]

Coincidentally I was working on a cam timing spreadsheet for another engine. I loaded your example to match the duration & timing relative to TDC & BDC. The only way I see it working is if the LSA (lobe separation angle) is 100-deg. LSA is represented on the chart by the line starting at 260 deg, ending at 460-deg which corresponds to EXH & INT cam centerlines respectively (their maximum lift points). To confirm, 460-260=200 deg CS angle. 200/2=100 deg between cam lobes.

I think it suggests LSA is equally important. For example if LSA is increased/decreased, the red/green EXH/INT bars diverge/converge accordingly. Even though the individual cam duration remains unchanged, the resultant timing angles relative to TDC/BDC will be different. For example, if one just assumed a right angle 90-deg LSA without knowing better, it would be physically impossible to achieve this particular timing given the stated cam durations. Does this math go around or am I missing something?

Indeed it does give you a 100 degree LSA. The thing is that LSA (like duration) is really a derived value- cam makers might like to quote it on the cam card, but from a design perspective it's better to decide when you want each valve event to happen, and let the LSA be whatever suits those requirements rather than the other way around.

One more thing I didn't mention - there is a tradeoff between cam lobe major diameter, valve lift and duration on the lobe, especially if you use flat tappets. Lifting too fast will result in the edge of the tappet catching on the lobe, which is sure to result in disaster. Roller lifters are less susceptible but can experience excessive side loads on the lifter bores if the ramp rate is too fast.

Sometimes you might find yourself using a less than ideal duration to get the amount of valve lift you need for adequate cylinder filling, particularly if your camshaft is small in diameter and you can't achieve a high rocker ratio.

 

[petertha]

One more thing I didn't mention - there is a tradeoff between cam lobe major diameter, valve lift and duration on the lobe, especially if you use flat tappets. Lifting too fast will result in the edge of the tappet catching on the lobe, which is sure to result in disaster. Roller lifters are less susceptible but can experience excessive side loads on the lifter bores if the ramp rate is too fast.

Another detail as others have mentioned is valve clearance effect, even in a perfectly tangency cam/lifter arrangement. This cam calls for 0.1mmm (~0.004") clearance which results in the geometric duration being reduced from 131 to 108 degrees ~17%. The smaller the cam is, the more that duration is reduced all things equal.

 

[Nerd1000]

Another detail as others have mentioned is valve clearance effect, even in a perfectly tangency cam/lifter arrangement. This cam calls for 0.1mmm (~0.004") clearance which results in the geometric duration being reduced from 131 to 108 degrees ~17%. The smaller the cam is, the more that duration is reduced all things equal.

Seems like a lot of clearance for such a small engine. Ideally the clearances would scale with the lobe size, but I guess the result would be impractically small clearances at model scale.

 

[petertha]

Actually not really, at least relative to manufacturer recommended setting of several commercial model engines I was able to check. Some are lower, some are higher, some differ between intake & exhaust. This is from an OS 4S engine & its quite prevalent across several engine types spanning different displacements, cylinder count & configurations.

 

[petertha]

I don't want to detract from Brian's post because he is after cam information for 'slow running' engines. But relating to general cam timing discussion, I did a summary of primarily commercial RC model engines link below. I've slowly been cataloging other shop made engines out of interest. That's what drove me to dig into this topic a bit deeper because some of the plans/instructions are bit mysterious to decipher, even with some homebrew computer tools.

https://www.homemodelenginemachinist.com/threads/cam-timing-methanol-glow-engines.33280/

 

[petertha]

and happens to be the default cam timing used by Lotus for their free engine simulation software.

Can you direct me to a link for this please?

 

[Nerd1000]

Can you direct me to a link for this please?

https://www.lotusengineering.com/engineering-software/

 

[Nerd1000]

Another detail as others have mentioned is valve clearance effect, even in a perfectly tangency cam/lifter arrangement. This cam calls for 0.1mmm (~0.004") clearance which results in the geometric duration being reduced from 131 to 108 degrees ~17%. The smaller the cam is, the more that duration is reduced all things equal.

This is actually really significant, the air going in and out of the engine only sees what happens at the valve. E.g. to implement the timing I suggested, with no overlap and ivc 20 degrees atdc, on the 20mm camshaft of my diesel design, I need a duration on the lobe of about 110 degrees due to the valve lash.

I should add that you can also achieve valve clearance by having parts of the cam shape be smaller than the base circle, or to put it another way the cam can be designed with 'negative lift' in the areas between the ramps. I imagine that modern car engines are usually done this way so they can have ideal dynamics, but this isn't going to matter for the likes of us.

 

[Vietti]

Obviously you can also, on an overhead valve engine, change the valve opening amount by changing the rocker ratios.

 

[petertha]

This is actually really significant, the air going in and out of the engine only sees what happens at the valve. E.g. to implement the timing I suggested, with no overlap and ivc 20 degrees atdc, on the 20mm camshaft of my diesel design, I need a duration on the lobe of about 110 degrees due to the valve lash.

Yes. That's why evaluating a cam profile in isolation of other contributing parameters can be misleading. It might explain why some model cams show overlap on a (theoretical) zero clearance, but once the true duration is calculated based on necessary clearance, the timing bars will separate from one another because they are both being reduced. Sometimes the feeler gauge setting is defined by engine designers, sometimes its not.

On a similar theme of incomplete information, here is another example. Just made up numbers but demonstrates the point. I'm show 2 quite different timing scenarios, but they are actually identical cams (open duration)& LSA. But the designer has not further defined an INT/EXH event relative to stroke position like TDC. So one scenario assumes INT open at TDC. The other scenario assumes INT open & EXH close equally spaced about TDC. The resultant timing 'numbers' relative to TDC/BDC are significantly different. This would be equivalent to shifting the camshaft over a couple cog teeth on the timing belt.

 

[peterl95124]

Brian,
is this slow enough for you ?
the secret is not cam timing, its the massive flywheel !
Peter.