Saturday, June 6, 2020

All Parties Heard From

Came in today to receive a nice email from the crankshaft vendor of choice, David at Marine Crankshaft, Inc.  He had reviewed my drawings and requests and agreed to take on the process of turning a rough-machined crankshaft blank into a ready-to-use part.

The plan is to use 5 1/2" diameter 4340 vacuum melt material from Yarde Metals in a normalized and tempered state (Rc28-34 for you techies) which will be machined in the lathe and mill, leaving adequate stock on all bearing surfaces.  The process will be very much like the process used for the V4 engine detailed here, here, here, here, and here, but with bigger, more rigid machine tools.  I'll then send the crank to Marine where they will heat treat, finish grind, detail oil holes, superfinish, then plasma nitride as the final step.  This will produce a crankshaft with extremely fine bearing journal finish, strong and ductile core section properties, harder/stronger shell properties, and an extremely hard and lubricious plasma nitrided final surface.  The last truly beautiful crank I saw was a Rick Schell Stage 4 crank for a TZ250.  It was so gorgeous you didn't want to put it inside crankcases.  I am hoping the Hypermono crank will elicit similar feelings.

The main dimensions of the crank (journals diameters, stroke) were retained from the Ducati part to make bearing sourcing a simple proposition but the other proportions were made to match the new single cylinder application.  The resulting design is short and rigid, two nice adjectives to use for a crankshaft.
I used as many of the design tricks for a high performance crankshaft as possible, including hollow rod journal, large bearing fillet radii, and tungsten slugs for balancing.

Determining the optimum engine balance factor was not straightforward.  The offset cylinder complicated the situation enough that a simple piston primary and secondary force analysis was not sufficient.  Again I called on the help of Tony Foale to determine the best approach.  One of his papers on engine balance  and his basic engine balance software were a good starting place.

Digging into his big bag of software tricks he was able to modify an existing program to take into account not only the basic piston primary and secondary forces, but the relative positions of the crank, piston axis, and counterbalance shaft.  He ran a few optimization studies and the results started flowing.  For a basic analysis a 50% balance factor on the crank and 50% on the counterbalance shaft produces the lowest overall engine vibration levels.  The optimized result was a few percent different than the basic simulation and, just as important, the balance weights are optimally not 180 from the throw, again off by a few degrees.  Even though 50/50 and 180/180 would have been good enough and resulted in a smooth engine, if I am going to go through all this trouble, why not make the parts to the optimal values instead of approximate ones?  There is no good answer not to!

The crank assembly is configured a little different than most crankshafts due to the idler gear being between the crankshaft and the clutch.  This gives me the same radial room on both ends of the crank for components.  As the design progressed and I shuffled components back and forth, the best overall layout ended up with the primary drive gear and generator on one end, and the timing wheel/starter clutch, starter gear, and cam drive sprocket on the other.

Most of these parts are Ducati OEM, no reason to reinvent the starter clutch or generator rotor.

Now that i have a definite path forward on all the outsourced components the last bits of material will be ordered, CAD files finalized and frozen, and toolpath generation started.  Lots of clicking ahead for me in the next several weeks but after that, fabrication starts in earnest.  That will produce much more exciting pictures and videos!

Friday, June 5, 2020

Details, Details, Details.

There seems to be plenty of empty time these days so there is plenty of time to obsess over some bike design details!  I'm sort of going a little stir crazy and am finding it hard to be really productive, so please bear with a scattershot blog post.  I'll be covering a bunch of stuff, from parts status to design details, to fabrication prep.

The last, last minute change is that we have decided to go from a 1299 piston/cylinder assembly which is a 116mm piston in a thin steel liner, to the 1199 piston cylinder assembly, which is a 112mm piston in a thicker aluminum liner.  Seems that the 1199 is the basis for all the race development, like 2 ring pistons for less frictional drag.  The lost 50cc of displacement will be more than made up for in less friction and better breathing.

Now that the important dimensions have now been set in stone, purchase orders have been issued and the bank account is cringing in submission:
  • custom length connecting rods from Carrillo 
  •   custom length high performance roller cam chain from IWIS
  • 6061-T6511 aluminum blocks for crankcases 

The engine CAD model is undergoing subtle refinements as time allows.  Current state is good enough to freeze and start making chips but there seems to be no sense rushing things so I am revisiting details that, while satisfactory from a pure engineering perspective, just don't have that elegant feel to the design solution.  These areas are, of course, few and far in between.  ;)

I have contracted with Curto-Ligonier Foundries to sand cast the two side covers and the oil sump in AZ91 magnesium.  Curto will be providing all the feeding, gating and sprue design information which I will use to create cope and drag parts of the molds for sand casting. 

The sand molds themselves will be 3D printed sand that the metal will be poured directly into.  The sand molds will be 3D printed by Humtown Products.  Brandon Lamoncha there has been a big help in gathering the various details needed for the process.

As the engine nears design completion, I can focus on the rest of the bike.  Since the chassis and suspension are already fully detailed, I can progress to details like air intake routing, airbox, and electrical components.

The dual air intake path threads through both small triangles formed at the crest of the frame then merges, has a filter mount, then enters the airbox.  Its volume is approximately 9.5l, plenty big to tune the Helmholtz resonance frequency to help boost the mid-range power.

For the air runners, I determined the minimum air intake size, doubled it, then used Creo's advanced 3d surfacing features to create duct with a curvature-continuous walls and a linearly increasing cross-sectional area. This helps slow the air from road speed to 0 with minimal losses, ensuring max static pressure.  Even though ram air benefit is pretty small for a bike that can only do about 160mph, it makes sense to try to get what you can if it does not get in the way of something else that is more important.

The air intake system leaves plenty of room for a nice size filter at the front of the airbox housing.  The various electronics modules (ECU, water pump control, coil, IMU, etc.) will live on to and around the base of the ram air ducts.  I am not sure if there will be brackets molded into the ducts or a complex 3D printed housing that has a few discrete mounting points.  Those details have a minor enough impact on the surronding parts that I can push them back until the first test assembly phase, where details like this are best resolved.

That's about it for now.  Forthcoming posts will cover the sand mold design process and the resultant castings, and the start of crankcase and crankshaft machining.

Hope you are all staying sane during this craziness.