Honda / Acura K20a K24a Engine Forum banner

21 - 28 of 28 Posts

·
Arouse the DAMPFHAMMER!
Joined
·
6,069 Posts
That's an awesome work Scider. I really appreciate your measurement!

It's somewhat interesting to see the offset between the lobe centers for high and low speed...This is probably designed around the factory VTEC switching point so the cam phaser doesn't have to move.
The setting of the centerline is mainly driven by mean VE and peak VE for each cam and the application. Of course it is also depended on the cam duration and profile shape. The lower the speed, the less gas dynamic support comes from impulse of the gas column, it's more related on the pressure swing. This is the main driver for VE and therefore the camshaft centerline. At higher speeds, even with higher duration of valve opening, the gas impulse and the the more retarded pressure wave helps to keep the air inside when piston is already moving up.

It makes me curious what aftermarket manufacturers are doing with their cam designs, and what design philosophy they use for the low speed cam profiles.
I know it exactly from TODA A3: low speed cam exhaust centerline is advanced by 4° compared to high speed cam. Centerlines of low speed and high speed are almost the same, as it is just a VTC work thing. A note, the low speed intake camshaftes have different centerlines by 2.5° difference, to start a sort of swirl for an faster combustion when the intake gas speed is low (part load - emission game). TODA overtakes that from the OEM camshafts.
 

·
Arouse the DAMPFHAMMER!
Joined
·
6,069 Posts
Ideally you want the flame to consume the end gas near the exhaust side before the intake, but this doesn't always happen based on port geometry and valve events.
Larry Widmer of ENDYN worked on that approach with his Roller Wave piston and head design. Knock location and activation has many parameters as you mentioned Scider. But this means it has also many options to depress it. These are the good news. The bad news all measures against knock cost something: emissions, efficiency or performance. Combustion design is finding the best compromise of those. One of my ideas lead to the highest compression ratio combustion systems in the market (Industry, CHP plants) while keeping emissions below limit at the highest available thermal efficiency of the market for such a combustion system. It's a different approach then Larry used for the Roller Wave concept, which is activation energy driven. My idea is mainly driven by turbulent kinetic energy. So two total different approaches for the same goal.

What OEM automotive combustion designs for down sized engines often miss is an idea to prevent knocking with more then the conventional primaries: retard ignition timing and adding fuel. There are so many options available and known, one of the OEM's I really appreciate is Mazda. They didn't follow that dead end strategy of down sizing and went into an high compression ratio design with all disadvantages and fought them successfully. The SkyActive gasoline engine is one of my favorite combustion designs in mass production engines.

Regarding fuel to cool turbine blades, the temperature level declines on both side of the optimal combustion velocity lambda. The right side is a much more difficult side, running lean means alot of effort, but it has an huge impact on efficiency too. I was involved in designing highly boosted lean burn combustion concepts and the main key to handle is ignition. Many OEM's have no experience in this field and even don't know technologies to handle it. The Corona Ignition System (Corona coming from the Sun-Moon-constellation not from the flue) is the wrong way into that. It's limited, expensive, and effectiveless to go really further. Formula One did the right step, pre-chamber combustion design. In industry this is common since the 80'ies 😂 . Just imagine, on a 27 bar BMEP engine having an exhaust temperature of 580 °C. This means alot to the boost concept and pressure drop over the engine. Finding positive pressure drops (= exhaust pressure - intake pressure < 0) concepts need some serious turbo technology. Anyway, running lean at almost a lambda of 2 and having a smaller (!!!) combustion duration compared to lambda 0.9... is included in a pre-chamber combustion design. I saw up to 20° less, just by shorting the combustion length. Faster is only HCCI, which is load limited.
 

·
Registered
Joined
·
81 Posts
Discussion Starter #24
Been putting all my effort recently into my Exocet build, and I suffer from over analyzing every minor aspect of it, so everything takes forever.

I spent some time trying to develop a more accurate lift curve for the rockers before I came to the conclusion my setup has too much slop and my gauges aren't sensitive enough at low lift values. I just couldn't find enough repeatability to be happy. It's something that's more reasonable to determine hypothetically from CAD geometry and doesn't have a huge impact to valve movement. Basically the ratio starts ~1.7 when the valve opens and increases to ~1.8 at stock VTEC lobe peak lift values. Aftermarket cams might approach a 1.85 ratio before the valve.

I do need to dig out the cam and valve spring measurements I took though. I wanted to investigate what sort of safety margin Honda is using for the factory valve springs with regards to over the nose cam acceleration and spring pressures on the VTEC cam. I only have the K24A2 cam and K24A2 springs to look at though. I measured a K20 cam as well, but I have no K20 example springs for the time being. The only other springs I have are the beehives from Ferrea (PN: S2018).

This is a bit off thread topic, but I've been compiling a number of dyno data sets into one spreadsheet and looking at what kinds of intake/exhaust/cam combinations they're using. A number of K24 dyno runs with aftermarket cams show decreased torque at low speeds while on the primary lobes compared to the stock cam, and this is something that's always bothered me, but nobody seems to pay much attention to it. VTEC lobe cam specifications are difficult to come by, and primary lobe data is almost non existent. It is well documented though that the K20 uses a relatively small primary lobe compared to the K24. The larger displacement makes the K24 less sensitive to a larger low speed cam so it works out in the end. As a result, If you were to put K20 cams in a K24 engine, you'd lose torque while on the primary cam, and the peak power might not improve enough to make up for it. My theory is some aftermarket cams follow the K20 design as a starting point, so when they're put on K24 motors, there is poor low speed performance and a large VTEC surge when jumping to the high cam.

For example, KMiata did a back-to-back with a K24 on stock cams to some DC 3.2 cams:


It's pretty obvious the stock cams deliver more torque below 70 mph, and you can see the switchover point for the 3.2 cams ~62 mph, where the stock cam switchover is much more subtle. VTEC surge might be an exhilarating kick in the pants feel, but a road race engine would be nicer to drive with more linear power delivery like the stock cam, so I've been trying to find aftermarket cam manufacturers that follow a K24 design philosophy. I personally like the look of dyno pulls i've seen of the 4Piston RR3 cam.
 

·
Registered
Joined
·
39 Posts
Nice to see another MEES student into K's! I am MEES class of '88, and have been in engine engineering for about 40 years. I have read through this thread, and while I applaud your efforts to design new cams, you are missing a very important aspect of how cam lobes are designed. I spent about a decade doing this and valvetrain development early in my career. When the OEM's design a cam lobe, after meeting all of the static and kinematic requirements (like peak contact stress, peak roller angle, spring margin, radius of curvature, etc) we then run the candidate lobe through a dynamic simulation program to see how that lobe excites the valvetrain and valve-spring. This is roughly proportional to the ratio of the acceleration pulse-width to the natural period of the valvetrain and valve-spring. The actual no-follow speed is always lower than the kinematic one due to vibrations of the valvetrain and the spring. This type of analysis is also critical for determining valve seating velocity. Too high a seating velocity wears out seats and can break exhaust valves. A good cam lobe design gives low excitation level to the valvetrain. There is some specialty software for doing all of this.
 

·
Arouse the DAMPFHAMMER!
Joined
·
6,069 Posts
...I personally like the look of dyno pulls i've seen of the 4Piston RR3 cam.
Would be glad to see it. Ben45 reported a while ago these cams of Webcams have steve ramp and high accelerations. It ate the TCT quite fast. He swapped back to his previous camshafts for reliability reason. Wouldn't interesting to see these measured.
 

·
Registered
Joined
·
81 Posts
Discussion Starter #27
Interesting where you run into people! Only one more class to go before the Class of 2020 graduates ;)

I've been doing calibration for about five years now, but the mechanical design is something I haven't had any exposure to outside my personal readings and the much more in-depth info I've gotten from the engine design courses in the MEES program. I was going through the course info regarding valvetrain analysis that you mention. I'm aware there is an entire field of dynamic study that I'm not going to be able to even begin to touch on. I suppose part of this whole journey is to practically apply lessons from coursework to something I'm passionate about, I've always found that the learning sticks better that way. I'm not sure I'll ever even get to cam profile design either, there's a whole mess of approaches and this really isn't the forum for that level of detail (although I'm partial to constant acceleration, biased by my grandfather's constant discussions of his days at Cosworth).

I'd say my end goal is to better understand valve train systems and camshaft profiles to the degree that I can, and to develop better tools to evaluate aftermarket parts. I'm skeptical of aftermarket suppliers, cams in particular due to the lack of any real meaningful information. They always seem to be paired with valve springs of ridiculously high seat and open pressures, I imagine more for to cover their ass from user error than anything.
 

·
Arouse the DAMPFHAMMER!
Joined
·
6,069 Posts
Nice to see another MEES student into K's! I am MEES class of '88, and have been in engine engineering for about 40 years. I have read through this thread, and while I applaud your efforts to design new cams, you are missing a very important aspect of how cam lobes are designed. I spent about a decade doing this and valvetrain development early in my career. When the OEM's design a cam lobe, after meeting all of the static and kinematic requirements (like peak contact stress, peak roller angle, spring margin, radius of curvature, etc) we then run the candidate lobe through a dynamic simulation program to see how that lobe excites the valvetrain and valve-spring. This is roughly proportional to the ratio of the acceleration pulse-width to the natural period of the valvetrain and valve-spring. The actual no-follow speed is always lower than the kinematic one due to vibrations of the valvetrain and the spring. This type of analysis is also critical for determining valve seating velocity. Too high a seating velocity wears out seats and can break exhaust valves. A good cam lobe design gives low excitation level to the valvetrain. There is some specialty software for doing all of this.
We did this also at my previous company, it's a good indicator for durability testing, chossing the right variation of lift, shape and duration to lower testing cost. But it doesn't replace it. Only the engine tells us it finally.

I designed valve lift curves for my DAMPFHAMMER engine, luckyly I found a aftermarket cam which was in near of it. Describing models for the dynamic motion of the K-series valvetrain are complex because of the roller and won't meet a good precision to condense down to the right cam lobe profile in a straight way.

In case of the RR3 profile I am not sure if final testing is transfered to customers. But hey its racing ;). Maybe Webcams did Spintron it, but even that is no guarantee when weights are different.
 
21 - 28 of 28 Posts
Top