Author Topic: Team Go Dog, Go! Modified Partial Streamliners  (Read 1438386 times)

0 Members and 7 Guests are viewing this topic.

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2580 on: October 09, 2016, 01:52:40 PM »
Rose does the books and knows exactly how much I spend, most of the time.  Lots and lots of money has been spent since the Pendine race with no results and she knows this, too.  She hasn't said anything negative and I am staying within my $500 a month race budget, sort of.  She rode pillion on our first date on the back of a BSA A-65 Spitfire.  That was 40 years ago.  Maybe she figures bikes and husband are a package deal.

The first stage in the build design is done and it used the filling-emptying model with ten point cam data.  There is a procedure for optimizing bore and stroke.  That cannot be changed due to budget considerations so I did not look at it.  The first thing looked at was inlet port minimum area.  The formula is on Page 250 in the users manual and it is (RPM x Stroke x Bore Squared) / 190,000.  The minimum area is 2 x (8,500 x 2.677 x 3.8**2) / 190,000 = 1.73 square inches.

The minimum port area is between the valve seat and the hole for the valve stem.  It is 1.89 square inches for the two ports.  The engine has adequate intake port minimum area.  There are all sorts of problems if this area is too big so it will be left as is.  The intake valve sizes will stay at 2mm larger than standard.  There is flow data for this size valve and an old and smaller manifold so I had to estimate flow with the new bigger manifold.  Kibblewhite is asked to do a flow test with the bigger manifold so I can rerun the simulations using measured values.

The second thing examined was the exhaust valve size.  There is a measured flow vs lift curve for 26mm valves.  The low lift flows for bigger valves are extrapolated based on increases in curtain periphery.  The high lift flows are based on the proportional increase in valve head area.  The flows at intermediate lifts are interpolated between these.  This represents bigger valves, seat diameter enlargement, and some grinding on the ports.

All cam timing simulations compared the three different valve sizes, 26, 27, and 28 mm.  The compression losses due to combustion chamber modifications were also considered.  This is 11.7 to 1 with standard exhaust valves, 11.5 to 1 with 1mm larger valves, and 11 to 1 with the 28mm valves.  The 27 mm exhaust valves consistently gave better performance in the simulations.  The flow data estimates are sent to Kibblewhite.  They will be asked if these flows appear to be obtainable with using the existing valve seats, enlarged a tad, with 27mm valves.  This will be done if it is possible.  The compression loss associated with bigger seats will be too much and the 26 mm valves will be the best option if this occurs.

Cam timing was the third thing considered.  Sensitivity analyses were made to see how performance changed with variations.  Lift was not a significant factor as long as it was over .370 to .380 inches.  Duration was not critical as long as it did not get too far from the optimized values of 273 for intake and 284 for exhaust at seat-to-seat (.006 lift).  Cam timing was most critical with 102 for intake centerline and 115 for exhaust.

The all-purpose cams I have used for years have .380 lift.  This is adequate.  Seat to seat duration extrapolated from .050 timing values is 291.6 for both.  These cams, with intake and exhaust lobe centers at 101.5 and 114 give good simulated horsepower.  These cams will be used.  Kibblewhite is asked to send me digital lobe profiles so I can do more complex wave simulation modeling.  They are also asked to check spring tension and valve train harmonics with these cams.

Last, rpm at peak power was looked at.  Cam timing and valve size could be varied to move the horsepower peak up from 8,500 rpm to 9,500 or higher.  Peak horsepower did not increase.  Torque decreased and this was offset by the higher rpm to produce the same power.  Peak power target rpm is kept at 8,500.  There is less chance of wearing out and breaking stuff at that rpm.

That is it for now.  The head, cams, etc were sent out yesterday.

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2581 on: October 15, 2016, 12:29:41 AM »
The simpler filling-and-emtying model assumes the intake and exhaust systems are optimized.  The more complex wave-action model is used to design these two systems.  A digital model of a cam lobe needs to be imported into the program.  It cannot use the simple 10-point cam descriptions.  The fellow who is going to digitize my cams is out of his office on business and I want to start on the modeling.  This is an imported Comp Cam profile that is similar to my cams if I enter a 1.063 rocker ratio.  Using this, I can rough in the intake and exhaust system designs.   

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2582 on: October 15, 2016, 11:07:47 AM »
Some wave analysis shows me that tuning the intake system is a no-brainer.  The intake length tuned to the second harmonic gives more power but with a peakier power spike.  The intake length for the third harmonic gives a bit less power but a wider peak.  Preliminary results say moving the bell mouth back about 3/4 of an inch should do the trick for max power at 8,500 rpm with the third.  This will be fine tuned when I get the lobe profile.

The filling-emptying model assumes the exhaust system is optimized and it shows a big power gain with a stepped header and collector system as compared to the one I have.  It is an "H" header with two glass pak mufflers.

The picture shows the race exhaust system on the Triumph factory team mile bike.  It is a stepped header going into a collector and it is basically what I need.  The header pipes are parallel for a considerable distance upstream from the collector so there is room for me to move the junction for fine-tuning.  The program will give me the info I need to do this.  It looks like I can get my dirty little paws on one of these.   

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2583 on: October 18, 2016, 12:21:35 AM »
Now it is time to start the virtual build.  A Comp Cam 9042 profile is selected with .360 lift, 265.8 duration at .006 (seat-to-seat), 221.1 duration at .050, .55 base circle radius, and default lobe centerline angles of 110 degrees.  It will be used for intake and exhaust.  The first pix shows its profile being imported into Dynomation.  The entry screen with the 110 lobe center angles is shown in the second pix.

The cam timing is optimized for the greatest area under the horsepower curve between 5,000 and 9,000 rpm.  265.6 intake duration and 277.2 exhaust duration are recommended.  The 9042 lobes are not far from this.  103.9 and 113.3 intake and exhaust lobe center angles are recommended.  The cam manager screen showing this is in the third pix.  The input screen for the cams set to those specs is show in the fourth pix.



   

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2584 on: October 18, 2016, 12:40:07 AM »
The 9042 cams produce 83.40 flywheel HP at 7,500 rpm with 71.69 pounds-foot torque at 4,500 rpm.  This is shown in the first pix.  The valve clearance I am using is .008 intake and .010 exhaust.  There is a good chance I can nip that up a couple of thou with no problems, and especially if the valve heads are thermal barrier coated so they stay cooler and the valves expand less.

The lash is nipped up to .006 intake and .008 exhaust.  The input screen in the second pix shows this.  The cam timing is optimized again and the lobe center angles are 102 and 113.4 for the intake and exhaust cams, respectively.  Power is now 86.59 at 8,000 rpm and torque is 71.85 at 4,500 rpm.  A few horsepower more and worth the trouble.  This is something I can discuss with the expert at Kibblewhite.

Now, I have the cam profile and it is Comp Cam 9042, the lobe centerline angles at 102 intake and 113.4 exhaust, and the lash at .006 intake and .008 exhaust. 


Offline Freud

  • Hero Member
  • *****
  • Posts: 5419
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2585 on: October 18, 2016, 01:05:30 AM »
 Duration was not critical as long as it did not get too far from the optimized values of 273 for intake and 284 for exhaust at seat-to-seat (.006 lift).  Cam timing was most critical with 102 for intake centerline and 115 for exhaust.


(.006 lift?)

Really

FREUD
Since '63

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2586 on: October 18, 2016, 09:33:28 AM »
The seat-to-seat valve timing is defined as .006 lift in a Society of Automotive Engineers guideline or standard.  This accounts for some variation in the base circle radii due to manufacturing tolerances.   A lot of lobes do not have a fully round base circle and there is one low spot on the base circle and everywhere else is some sort of lift, even if it is a few thousandths.  The seat to seat duration would almost be 360 degrees when measured with no tappet clearance.  This way of defining seat-to-seat allows for small imperfections in the base circle radii.

It is a controversial practice.  More than one cam grinder tells me that it means nothing as per comparing cams against each other in real-world performance.  They say I should only look at the part of the lift curve when the valve is open enough to provide flow.

The software uses these .006 lift figures for all sorts of things.  My feeling is "when in Rome act like a Roman" as the saying.  So, I do things the way the user's manual and tech support guy says.   

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2587 on: October 19, 2016, 12:56:55 AM »
The intake tract length from valve face to the primary expansion point, the minimum intake port area, and the intake area just inside the intake bell is entered.  The minimum exhaust port area is entered along with some initial exhaust system specs.  There are formulae in the users manual to compute the latter.  Now it is time for wave analysis.

The intake tract length is optimized first.  The primary expansion point can be moved about 1/2 an inch back or an inch forward.  These are the screens I am looking at to do this.  The vertical striped bar on the top graph shows the rpm that is being analysed.  HP and torque are shown, too.  The second graph shows port pressure with the green line being the intake and red line the exhaust port.  The space between the vertical lines IVO on the left and EVC on the right are the area of concern.  This is the overlap period.

Cylinder pressure is shown on the bottom graph at two different scales.  The green is an overview of pressure throughout the entire cycle and the red is a detail of the pressure during overlap.

The goal is to have the pressure in the intake tract drop during the overlap.  The green line slants down from left to right between IVO and EVC when this happens.    Fresh charge is pushed into the combustion chamber when this occurs.  Note that at 6,000 rpm there is very little pressure drop.  Things are a lot better at 8,000 rpm.  There is a big pressure drop during overlap. 

Various intake tract lengths are looked at within the 1.5 inch adjustment range and the best one will be selected.


Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2588 on: October 20, 2016, 09:53:23 PM »
Within the overlap period the inlet port pressure should decrease as the crank angle increases.  Five intake tract lengths are examined, 9.2, 9.7, 10.2, 10.7, and 11.2 inches.  The port pressures had the best gradient with a 10.2 inch inlet tract and this is a half inch shorter than it is now.  Here are the printouts from 5,000 to 9,000 rpm. 

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2589 on: October 20, 2016, 09:55:23 PM »
And 9000 rpm.  The horsepower numbers also show this length to be the best.  Next, the exhaust system will be optimized.

Offline Koncretekid

  • Hero Member
  • *****
  • Posts: 1203
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2590 on: October 22, 2016, 09:56:24 AM »
"Duration was not critical as long as it did not get too far from the optimized values of 273 for intake and 284 for exhaust at seat-to-seat (.006 lift)."

A cam with durations of 273/284º at .006" lift would only have durations around 220-240º (or less) at .040" lift.  Isn't this a very mild cam?  I don't have any experience with OHC 4 valve motors, so maybe they don't need the long duration valve timing that our pushrod 2 valve motors.

Tom
We get too soon oldt, and too late schmart!
Life's uncertain - eat dessert first!

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2591 on: October 22, 2016, 12:04:59 PM »
The two lobe profiles I can find are for Comp Cams grinds and they are supplied with the program.  The best one is this 9042 profile.  There is no success in getting profiles from anyone who grinds these cams.  There are lobe profiles for radical full race cams I have.  They were digitized by Kibblewhite.  The intake cam would beat the valve gear apart pretty quick.  The exhaust cam is OK in that respect and I am modeling it in an alternative virtual build.  The old all-purpose cams I have been using for years have decent lobe shapes.  Kibblewhite is digitizing them.  The combos I am looking at in the other model are old intake and old exhaust, new exhaust used as new intake with old exhaust, old intake with new exhaust, and two new exhausts used together.  The folks that sold me these want to keep their details off the net so you'se guys see the 9042 cams in the build diary.  They are limp, no doubt.

The interpretation of squiggly lines is too much for an already confused mind.  The old guy understands horsepower.  Possible intake tract lengths with the range of adjustment are analyzed at half inch intervals.  HP at various rpm are tabulated.  See attached table with highest HP numbers circled.  10.2 inches looks best and it is a half inch shorter than it is now.  Note the high HP sneaking onto the bottom row of the right column.  That is another harmonic.  It would be the one I would tune for if operating rpm was more than 8,500.

The users manual gives a formula for minimum intake port area.  The ports are OK based on it.  They recommend looking at mach number, too.  Recommendations are .5 to .6 at peak HP for an engine built for top end power.  The far left column in the printed table is crank angle on a 720 degree basis and this is at 9,000 rpm.  The mach number is in the column to the far right and it is at the lower end of the recommended values.  The intake port throats are plenty big and nothing good will happen if they are enlarged.     

This is consistent with guidance written by Vizard.  He mentions a 1mm increase in valve sizes being enough for a four valve head.  A 1mm valve size increase on all four with a good port job would be ultimate for racing.  This head will have 2mm oversize intakes with 1mm bigger exhausts.  The only reason for the 2mm bigger intakes is the need for a bit extra flow at very high rpm and this is not needed for other types of racing.   

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2592 on: October 23, 2016, 03:30:13 PM »
The intake tract length is entered as 10.2 inches.  This is the best based on the prior worksheet.  Packaging considerations limit the range of header pipe lengths from 23 to 33 inches.  A length of 25 inches works best.  The horsepower ratings are not very sensitive to header length.
 

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2593 on: October 23, 2016, 03:36:59 PM »
A 25-inch header length is used when the header step dimensions are optimized.  An 1.75 inch initial diameter and a 1.5 inch final diameter work best.  Packaging problems make the 1.75 inch tubes awkward. The initial diameter is reduced to 1.625 inches.  This is easier to fit and there is minimal sacrifice in power.  The exhaust dynamics are sensitive to step dimensions.

Offline wobblywalrus

  • Hero Member
  • *****
  • Posts: 5503
Re: Team Go Dog, Go! Modified Partial Streamliners
« Reply #2594 on: October 23, 2016, 03:42:13 PM »
The 1.625 - 2.0 step diameters are entered.  Collector length is calculated next.  A 12 inch length is selected.  This gives decent power and the end is in a good location, packagewise.  The power ratings are not sensitive to collector length.