Milwaukee Midget

[Old Scrambler]

Midget and F-B.............I've been following your thread for a few years...............and picked up several pointers........thanks.......and congrats on the record achievement. I live a little deeper into the woods north of the chedder curtain. I also took a late-life interest in competing on the salt. My motto is 'going fast - slowly'

So last year Dave Murre, my engine builder and all-around fabricator / race adviser, helped me assemble a HONDA CB750 motor to compete in the BMST event under the Classics category - 1955 to 1980. The photo shows my custom designed valve-train incorporating modern valves and springs with a ported and bored head. I began with the idea that if Honda made bikes and cars in the same plant during 1966, they just might use the same sized castings to hold the valve-guides in their modern car motors.............and YES, they do. That meant the guides snapped into place and the valves fit without anything more than a mild hone/ream.

I searched the FERREA site and ordered a trial set of intake and exhaust valves with all components including bee-hive springs and ti-retainers. It all fit except for the stem-length which was purposely ordered long to allow Dave to machine to an exact length and cut new collet-groves. That allowed us to get the desired spring pressure.

The FERREA valves use a very steep seat-angle which allowed (begged) us to bore the valve-throats  and use narrow seat surfaces. The finished product moved the HP and RPM range significantly upward. We have made a few trial-runs but did not have enough time to get the timing and carbs properly tuned in 2014. We have since completed a dyno-test and did several pulls at 11,300 rpms with no apparent motor damage. The oil and sump were clean.    

I'm passing this info to you for consideration as you look for 'go-fast' motor components on a budget. BTW......Millenneum Technologies in Plymouth is the local dealer for FERREA.  

[fordboy628]

Scrambler,

What would be interesting to know about your "build spec" would be:

A)   Cylinder head flow values, both before, and after modifications.
2)   Cam profile (not just cam card specs), AND, cam timing(s).
d)   Anything else (such as C/R) that changed between builds.
z)   The balance of the "build spec".

I think everybody would be understanding if you wanted to keep some (or all) of your information confidential.

In general, valve/valve seat angles steeper than 45 degrees favor mid and high lift flows, at the expense of low lift flows.   Valve/valve seat angles less than 45 degrees (say 30 degrees, as on some older Pontiacs) favor low lift flows, but typically hurt mid range flow.    The reason is about the trigonometry of the opening (curtain area) at a particular valve/seat angle.   Lower angles have a larger effective opening at low lifts, due to the trigonometry.  At some lift point at or above the value where "curtain area = valve area", it ceases to have a large effect on the flow.   The "trick" here is to "maximize" the area under the flow curve, in a way that the engine is able to utilize it, for the intended useable rpm range.    It is really easy (with well designed heads) to present large flow values to the engine, which might turn your engine into a "high rpm winder", unintentionally.    Flow available needs to be "matched" to "flow demand" for the best torque and efficiency.    Easy to say, challenging to accomplish . . . . . .
 
My experience is that one variable of a "build spec" will not usually turn an engine into a "high rpm winder".   It's usually a combination of factors/variables.    This can usually be avoided at the design phase of the build, but only if someone takes the time to evaluate the components and the "build spec".

And, BTW, shallow valve seat angles typically do not seal well at high rpm's and elevated temps.    The original reason valve seat angles were made steeper (as far as I am aware) was to enhance valve sealing at high rpm's and during races of long duration, say 500 miles.    Enhanced valve sealing translates into more dynamic pressure in the cylinder . . . . . . etc, etc, etc.

 
Fordboy

[Koncretekid]

Scrambler,

   The "trick" here is to "maximize" the area under the flow curve, in a way that the engine is able to utilize it, for the intended useable rpm range.   


Fordboy

Do we really care about low lift flow?
Do we really care about mid lift flow?
Do we really care about maximizing  the area under the curve except for the last 500 rpm?

My experience is that my bike gets up to within a few hundred rpm of its max hp rpm and then takes a long time to get over the wall.  I think I'm only concerned about high lift, max hp rpm flow.  At my speed vs. drag curve, 1 more hp results in 1 more mph.  I don't care how steep the power curve is as long as the last 500 rpm give me the hp to speed increment I need.

For example, on my recent trip to Loring, I was going 140 mph at the 1 mile at 6400 rpm and 151 at the 1-1/2 mile at 6900 rpm.  I have approximately 50 hp at 6900 rpm, but I only need 40 hp to achieve 140. So it seems the really important part of the curve is that from 6400 rpm to 6900 rpm, i.e. the last 8% of the hp/rpm curve.

Which is maybe just another way to say "in a way that the engine is able to utilize it, for the intended useable rpm range."

Tom

[fordboy628]

Scrambler,

   The "trick" here is to "maximize" the area under the flow curve, in a way that the engine is able to utilize it, for the intended useable rpm range.   


Fordboy

Do we really care about low lift flow?      Only somewhat, but it is important for other applications.
Do we really care about mid lift flow?      Yes   
Do we really care about maximizing  the area under the curve except for the last 500 rpm?    Yes, but you need to optimize for your intended rpm range and number of trans gears.

My experience is that my bike gets up to within a few hundred rpm of its max hp rpm and then takes a long time to get over the wall.  I think I'm only concerned about high lift, max hp rpm flow.  At my speed vs. drag curve, 1 more hp results in 1 more mph.  I don't care how steep the power curve is as long as the last 500 rpm give me the hp to speed increment I need.

For example, on my recent trip to Loring, I was going 140 mph at the 1 mile at 6400 rpm and 151 at the 1-1/2 mile at 6900 rpm.  I have approximately 50 hp at 6900 rpm, but I only need 40 hp to achieve 140. So it seems the really important part of the curve is that from 6400 rpm to 6900 rpm, i.e. the last 8% of the hp/rpm curve.

Which is maybe just another way to say "in a way that the engine is able to utilize it, for the intended useable rpm range."    YES

Tom

Tom,

See my comments in your text.    I agree with your assessment, with one limitation.    As long as you have enough gears to keep the engine in a productive part of the rpm band, you will be OK, up to the point where drag hp = bhp available.    In 2013, using the close ratio 4 speed trans, once the Milwaukee Midget got into 4th gear, it would only pull up to the rpm where there was a torque & bhp "dip" in the curves, it would never pull past that spot.   I believe that the rear wheel torque available at that point was equal to the drag force for that speed, 118+ mph.    When the engine was redone for 2014, we worked hard to increase the bhp available and remove the "dips" from the curve.    That and other changes, worked out well.

Narrow and steep power curves work, as in F1, because they have lots of gears to keep the engine in the intended rpm range.   When your engine "falls off the power curve" it might not be able to pull itself back up into the peak part of the power band in the higher gears.

 
Fordboy

[fordboy628]

Okay, back to the more mundane (non beer-centric,   ) aspects of this thread . . . .

Here are the photos of the sealing gasket for the flow testing slider plate.    I bought a roll of 1/16th closed cell neoprene gasket material for this application, because the head gasket that was removed from the engine during disassembly was trashed.

The material is not as firm as some of the material supplied by BRP, but I was concerned that my slider plate, made from MDF, would be significantly less stiff than the BRP aluminum slider plates.    The combination performed well (0 leakage) on the flow bench 8-11-15.

   

The material is cheap, available from McMaster-Carr, and easy to cut.  I used a Harbor Freight gasket punch, a utility knife, scissors and a metal straightedge to cut it out.   It compresses to about half its' thickness under moderate clamping pressure.    It is slowly regaining its' original thickness, 2 days after being clamped up.    I think if I was going to be testing a lot of heads in one day, I might opt for the harder material like BRP supplies.

 
Fordboy

[fordboy628]

Some photos of the setup I used to rout the slider plate mounting slots.    It was a fairly complicated setup due to the proximity of the slot to the edge of the MDF plate.









The slot is actually 2 parts, a 5/16ths clearance slot for the 1/4-20 retention bolt, and a 9/16ths clearance slot for the tapped retention plate.    Both slots, on both sides, need to be in proper alignment to the slider mounting fixture.   It is important that both the retention bolt and plate end up slightly lower than the top surface of the slider plate.   This prevents any interference with the head, manifolds or radius adaptors.

 
Fordboy

[fordboy628]

Here are a couple of photos of the adaptor I made to open both valves at the same time.    I also made up some length adaptors to insert in between the indicator and the twin valve adaptor.    The "extenders" were threaded 4-48 and I made up .5", 1.0", 1.5", and 2.0" lengths in 3/16ths OD brass rod.     I used a 4-48 setscrew for the male thread portion.    The valve actuator rods are 10-32 threaded rods.    One rod is fixed and the other is adjustable, to compensate for slight differences in the valve stem lengths.    Fortunately, the valve center to center lengths are equal between the intakes and the exhausts, so only one "spreader bar" was required.    My design is simple and effective, but requires different "spreader bars" for differing valve center to center lengths.    BRP sells an adjustable spreader bar for this purpose.    My spreader bar is made from 1/4" x 1/2" mild steel of the appropriate length.    It attaches to the AGD indicator with a 4-48 x .5" SHCS.   (Socket Head Cap Screw)    Although I plan to only test up to .50" lift, a 1" travel dial indicator allows some margin for flexibility with the mounting bracket.

   

The indicator needs to be centered between the valve stems to actuate properly.    I also ordered some lighter "dummy springs" so that the indicator mounting bracket would not be forced into a "flex".    They are about 2/3rds of the tension of the dummy springs I use for 2 valve engines.

 
Fordboy

[fordboy628]

And finally . . . . . .

Some photos of the plate being fitted to the BRP fixture and some other photos of the head with inlet and exhaust adaptors for the flow testing.


Slider plate being marked for cylinder centers on BRP fixture.



Intake side being tested on cyl #1.    Note that the inlet radius adaptor completely covers the "slider slot".

The vertical lines on the slider block are the cylinder center lines, for alignment with the 3" diameter cylinder barrel mounted in the fixture.


Exhaust side test, cyl #2.    Slider assembly with head was "flipped" so that the indicator face would be visible and accessible.   Small PITA.

All things considered I was pretty happy with the way the adaptors worked out.   No problems and we were able to conveniently flow test every cylinder.


Next up, as soon as I fire up the portable cauldron, print out the data and scan it to PhotoBucket . . . . .

The flow bench results, AND, a comparison to the cylinder head on the venerable "Grenade".  (Maybe I should wait until "Stupid Joke Friday" to post the comparison data . . . . .   )

 
Fordboy

[Old Scrambler]

I'm with Tom (K-K) about only being concerned with the upper range of HP and finding the corresponding gears to KEEP the motor in that range at the critical top-gear change. No flow data here........Jerry Branch and others have flowed these motors about as many times as any other.........and more than most.  We started with about 60-bhp and moved up to a little over 80 at the rear wheel. The only printed dyno sheet showed 78 but the battery connection fell off

The custom-grind cam has wider lobe-centers which we believe move the HP curve upward in the rpm-range. OEM was at 8,000 to 8,500.........now its at 10,300 or so.  We timed the cam according to spec and then played with ignition timing and found significant power at a +9-degrees from spec advance.  The motor has a very rapid power-climb from 6500-rpms up in 3rd gear...............but our goal is to keep it above 8,000 when engaging 4th and 5th gears..............that means shifting at more than 10,000 rpms..........or PEAK HP.........and hoping the motor will pull better than 9000 in top gear.

I'll be travelling for a few days.............

[fordboy628]

I'm with Tom (K-K) about only being concerned with the upper range of HP and finding the corresponding gears to KEEP the motor in that range at the critical top-gear change. No flow data here........Jerry Branch and others have flowed these motors about as many times as any other.........and more than most.  We started with about 60-bhp and moved up to a little over 80 at the rear wheel. The only printed dyno sheet showed 78 but the battery connection fell off

The custom-grind cam has wider lobe-centers which we believe move the HP curve upward in the rpm-range. OEM was at 8,000 to 8,500.........now its at 10,300 or so.  We timed the cam according to spec and then played with ignition timing and found significant power at a +9-degrees from spec advance.  The motor has a very rapid power-climb from 6500-rpms up in 3rd gear...............but our goal is to keep it above 8,000 when engaging 4th and 5th gears..............that means shifting at more than 10,000 rpms..........or PEAK HP.........and hoping the motor will pull better than 9000 in top gear.

I'll be travelling for a few days.............

I need to know more about your engine type and build spec to comment intelligently.

Normally, widening the cam's lobe centers DOES NOT move the bhp curve upwards.    BUT, on a 2 valve hemi, (like the Triumphs I am familiar with), widening the LCA would provide more valve to valve clearance at overlap.    This might be done with a cam profile that has more "lobe area", in order to fit in as much lift and area as possible.    It would be the increase in lobe area that would raise the rpm level.

Like I said I would need to know more specifics . . . . . .

 
Fordboy

[fordboy628]

Rover K  flow bench testing results . . . . . . . .   Brought to you by Stupid Joke Friday . . . . . .

OK, as promised, the fruit of the flow testing . . . . . . .

I've arranged the photos and comments so readers can change their screen zoom level to read the graphs & data.
150% to 200% should work, depending on your eyesight . . . . . .    

The files:
Rover stock 01:    Base head totally stock
Rover stock 02:    Base head, stock valves back cut with 30 degree angle
03b (BMC):          Longman based BMC head with race valves, currently in use on the "Grenade".

Graphs are based on average tested flows for each head.

Comparison graph of corrected flow, intake:

Well, 4 valves are better than 2.

Data for intake graph:


Comparison graph of corrected flow, exhaust:

Again, 4 valves are better than 2.

Data for exhaust graph:


Percentage difference data for comparison between Rover 01 and Rover 02:

Data shows that a modest valve reshaping from the stock shape gains significant low/mid lift flow on the exhaust only.   Probably due to the fact that the stock intake valve had a modest back cut to begin with.

Percentage difference data for comparison between Rover 01 and BMC 03b:

I don't think any comment is required here . . . . . .    except, that I chose Stupid Joke Friday to put this up . . . . . . . .   

So we now have a starting point for some simulations.   More flow testing to follow once Chris makes some decisions about valves, etc.    Stay tuned.

I'm going to spend some time hand entering some flow data from the 'K' engine page.   Dave has flow tested and posted the data for some of the heads he has ported.    The comparison(s) might prove interesting.   It will be a bit until I get that finished up and posted.    Back to being Droneboy workin' for the Red Queen . . . . . . .   

 
Droneboy

[Koncretekid]

Scrambler,

   The "trick" here is to "maximize" the area under the flow curve, in a way that the engine is able to utilize it, for the intended useable rpm range.   


Fordboy

Do we really care about low lift flow?      Only somewhat, but it is important for other applications.
Do we really care about mid lift flow?      Yes   
Do we really care about maximizing  the area under the curve except for the last 500 rpm?    Yes, but you need to optimize for your intended rpm range and number of trans gears.

My experience is that my bike gets up to within a few hundred rpm of its max hp rpm and then takes a long time to get over the wall.  I think I'm only concerned about high lift, max hp rpm flow.  At my speed vs. drag curve, 1 more hp results in 1 more mph.  I don't care how steep the power curve is as long as the last 500 rpm give me the hp to speed increment I need.

For example, on my recent trip to Loring, I was going 140 mph at the 1 mile at 6400 rpm and 151 at the 1-1/2 mile at 6900 rpm.  I have approximately 50 hp at 6900 rpm, but I only need 40 hp to achieve 140. So it seems the really important part of the curve is that from 6400 rpm to 6900 rpm, i.e. the last 8% of the hp/rpm curve.

Which is maybe just another way to say "in a way that the engine is able to utilize it, for the intended useable rpm range."    YES

Tom

Tom,

See my comments in your text.    I agree with your assessment, with one limitation.    As long as you have enough gears to keep the engine in a productive part of the rpm band, you will be OK, up to the point where drag hp = bhp available.    In 2013, using the close ratio 4 speed trans, once the Milwaukee Midget got into 4th gear, it would only pull up to the rpm where there was a torque & bhp "dip" in the curves, it would never pull past that spot.   I believe that the rear wheel torque available at that point was equal to the drag force for that speed, 118+ mph.    When the engine was redone for 2014, we worked hard to increase the bhp available and remove the "dips" from the curve.    That and other changes, worked out well.

Narrow and steep power curves work, as in F1, because they have lots of gears to keep the engine in the intended rpm range.   When your engine "falls off the power curve" it might not be able to pull itself back up into the peak part of the power band in the higher gears.

 
Fordboy

Actually Mark, I didn't have my brain in gear when I wrote that post!  I was talking flow vs. RPM while you were talking flow vs. lift.  I agree that the area under the flow vs. lift curve is always important, but most important at high RPM (if we were actually able to measure that).  In any case, the cam in my motor is a road racing cam, as I don't know if anything more appropriate is even available, and my power curve looks more like a diesel truck, with a very wide flat hp curve so I don't have any appreciable dip to overcome until I hit the wall, right at my max hp vs. rpm intersection.  I'm going to try overgearing it this year to see what happens.

Incidentally, your flow bench testing is fascinating.

Tom

[fordboy628]

Actually Mark, I didn't have my brain in gear when I wrote that post!  I was talking flow vs. RPM while you were talking flow vs. lift.  I agree that the area under the flow vs. lift curve is always important, but most important at high RPM (if we were actually able to measure that).  In any case, the cam in my motor is a road racing cam, as I don't know if anything more appropriate is even available, and my power curve looks more like a diesel truck, with a very wide flat hp curve so I don't have any appreciable dip to overcome until I hit the wall, right at my max hp vs. rpm intersection.  I'm going to try overgearing it this year to see what happens.

Incidentally, your flow bench testing is fascinating.

Tom

Tom,

If your bhp curve is very wide & flat, there is probably something you can do to punch it up, BUT, it will narrow it as well.    Sometimes there are mechanical limitations of the engine's design ("packaging problems" . . . . or component strength) that prevent the kind of modifications that would increase power.

On a lot of the old style 2 valve hemi motorcycle engines, the limits usually are: valve to valve clearance, air cooled compression ratio limitations, and rocker ratios are also typically low.    The whole valve train has harmonics problems if I'm remembering correctly.

Thanks about the flow testing.    It's a lot of work, but essential to well engineered result.

 
Fordboy

[fordboy628]

Well, I spent some time hand entering the data that Dave Andrews posted on his Rover K engine page, into the flow testing and analysis program I use, Port Flow Analyzer.

Unfortunately, the data he has generated does not match up well with the data I have for the MM's new head.

So, kinda seems like a wank.

I will probably post some of it anyway, just to illustrate the differences that can exist between flow benches, AND, flow test procedures, as his procedures and adaptors (if any ?) are unknown.    One of the problems is that all his testing was done at 10" of test pressure Vs the 28" test pressure I prefer and use.

 
Fordboy

[fordboy628]

I meant to do this before, I just ran out of time.

Some photos of the stock rover valve shape Vs. the valves I back cut.


The intakes had a modest back cut originally.   The larger back cut really didn't do much.



The exhausts did not have a back cut to begin with and the low and mid lift flows picked up quite a bit from that simple modification.



It will be interesting to see the effects of reshaping the valve seat and pocket.     Hope to have a damaged head to experiment with at some point.    I also want to reshape the "choke point" in the port and reshape the guide boss and guide "nose".     Hope to also get some "race" valves tested back to back.

Ahhh, for those who might be wondering, that is a Huber Bock in the background.    Thanks to my 5' 17" buddy.    A tasty beverage on a hot summer day for "bingeneering" . . . . . . . . .

 
Fordboy