A Black Widow ... maybe

[mayhugh1]

I purchased a set of Black Widow castings from Dwight Giles during the 2012 Portland Gears Model Show. The documentation was a disaster, and after several phone conversations with Dwight I realized I'd need a lot more engine building experience before even thinking about tackling it. Still in their original shipping box, the castings were moved into my attic and in the intervening years twice offered for sale with no takers.

After finishing the 289 HiPo I began looking for another challenging project, and the castings were hauled down from the attic. Initially I wasn't enthusiastic about starting another V8 build, but the blower would be a new twist and something I could get interested in. I'm not sure how functional it was intended to be, but I'd already begun thinking about how I'd tackle a set of close fitting rotors. Ironically my documentation doesn't include any information on the blower.

Looking ahead, I'll likely include an electric starter similar to the one on the 289 which has been working well. I'm not yet sure about carburetion. There's nothing about a carburetor in my documentation, but a large colored poster that was in the box with the castings shows the engine sporting a pair of mystery Walbros. Also, with practically no documentation on the oil or water pumps, I haven't yet decided if there will be an oil pump, but the engine will definitely be water cooled. The camshaft will probably be a roller with the 289's profile, but with a slightly earlier exhaust valve opening. The immediate goal however is to focus on a foundational assembly of the block, heads, and intake manifold. With this in hand, a crank and camshaft will be designed to fit.

Chip making started with the rough machining of the block, head, and manifold castings to create a starting set of dimensions for their machined surfaces and features. Beginning with the block's pan surface, the absolute minimum amount of material was removed from each casting to obtain true and symmetrical mating surfaces. The 13 year-old age-hardened castings machined very nicely with only minor porosities showing up on the block decks. Dimensions of the oversize machined castings were carefully taken and used to construct SolidWorks (2010) models which were in turn used to create a virtual assembly.

Using the supplied drawings as a starting point, the cylinders and lifter bores and head bolts were located with respect to centerlines positioned on the oversized models. Centerline referencing allowed the features to be located on the models before they were brought to their finished sizes. The row of lifter bores which are immediately adjacent to the block's un-machined inner deck walls were used to construct the longitudinal centerlines in each bank to locate one axis of the drilled features. For the second axis a single orthogonal centerline for both banks was constructed in the middle of the un-machined web for the center main bearing. This centerline located the cylinder bores, lifter bores and the head bolts in both banks with the port bank features offset .156" behind it, and the starboard bank features offset .156" in front of it. The cylinder bores in each bank are 1.625" apart, and there is a .3125" offset between banks. There's .3125" between the lifters in each cylinder's pair, and each pair is centered over its cylinder.

A front cross-sectional view of the virtual assembly was monitored to ensure a vertical cam-crank separation of 1.405" inches to guarantee the engine's timing gear spacing. In my case it showed there was still .3425" excess material to be removed from the block's pan surface. Compared with the other surfaces this was a lot, and so it will be put off until the bore for the camshaft is drilled. This view also verified the locations of cylinder and lifter bore axes which must intersect the axes of the crankshaft and camshaft bores, respectively.

According to my documentation, a single head with identically located features was expected to be used on both banks. In my case the offset features between banks caused one of these heads to come up short of the block's front surface and leave a tiny gap. To avoid this I created unique port and starboard heads whose feature are offset from each other by .125". Hopefully this won't create problems with the yet to be located fuel, water and oil passages. They'll be modeled before any drilling is done.

A virtual model of the intake manifold was also created using the dimensions of the actual oversized machined manifold. Of the four castings, the intake manifold was the most difficult to accurately machine. Its ends were machined parallel to each other before its bottom and top surfaces were faced parallel. The two 45 degree surfaces were the most critical and were done in a couple iterations using small cuts. The virtual intake was spaced off the block and heads by .030" to account for the head and manifold gaskets.

Dowel pins were then added to the virtual block and heads and then machined into the actual castings. Both heads have identical dowel locations which may prove to be a mistake, but I thought doing so would simplify fixturing later on. A set of .030" shims added to the actual assembly allowed a continual comparison between it and the virtual assembly as it was being finish machined. The decks were left with .005" excess stock for a clean-up pass after the liners are installed.

For a final sanity check, the block was set up in the mill and each bank indicated on its dowel pins. A quarter inch test bar in the spindle verified the (.015") clearances between the un-machined inner walls of the block and the locations where the lifters are to be bored.

The next steps will include the final machining of the block's pan surface so the 7" deep camshaft bore can be drilled. The oil and water holes will also be located before any of the other features are drilled. - Terry

 

[GreenTwin]

Looks like a tricky build from many standpoints, as some have discussed previously.
Hats off to you for attempting this project.
I am ready with unlimited amounts of popcorn, and will be following.

.

 

[gbritnell]

I'm sure a lot of people will be following along to see what you come up with for dimensions.

 

[Ghosty]

Another one to watch. Should be interesting
Cheers
Andrew

 

[ShopShoe]

Terry,

Thank You in advance for starting what will probably be another outstanding series of posts on this build.

I always am amazed and learn a lot about problem solving from your projects, and this one will probably provide a remarkable number of examples.

--ShopShoe

 

[Cessnadriver]

Terry
There are many of us out there that purchased these castings myself included. I appreciate you taking on this project as it will be my guidance to making an engine less of a huge project.

Thank you and look forward to another of one your remarkable engine builds. If you are willing to sell the drawings at the end I know I would gladly pay for them.

Mark

 

[GreenTwin]

Sometimes you can trade off a little bore to get more design/working space, and accommodate casting variations.
An engine will still work well with a slightly reduced bore, assuming everything else gets worked out correctly.
Perhaps I am stating the obvious.
.

 

[k2steve]

Terry
There are many of us out there that purchased these castings myself included. I appreciate you taking on this project as it will be my guidance to making an engine less of a huge project.

Thank you and look forward to another of one your remarkable engine builds. If you are willing to sell the drawings at the end I know I would gladly pay for them.

Mark

Please put me on the list - I am not a design engineer and too lazy to figure it out on my own.

 

[mayhugh1]

Lots of drilling and a narrowly dodged bullet ...

Block preparation continued with the removal of a large amount of excess stock remaining on the pan surface. The heights of the deck dowel pin holes above the pan surface were used to reference the operation. A combination of slightly different draft angles and/or fillets on the outer walls adjacent to the pan surface makes the final result look irritatingly asymmetrical.

After completion, the block was set up in the mill for machining its .625" diameter 7" deep camshaft bore. Drilling the entire hole from one side of the block risked ending up with a hole that drifted off course. My experience with similarly drilled holes in billet blocks has been on the order of .005". Drilling from both sides however risks misalignment of the integral bearings inside the block. The cast-in openings between the main bearing webs adds risk due to the interrupted cuts through what will likely be nasty corner material. I decided upon a full depth hole drilled from a single side.

The bore's location above the finished pan surface is tied to the timing gear spacing (1.407") and was easy to locate on either end of the block. Perpendicular lines dropped from the edges of the machined deck surfaces to the edge of the pan surface were used to center the bore between the decks. These constructions were best done on the block's rear surface because of the mixture of machined and un-machined surfaces on its front end. As a result, the hole was drilled from the rear surface of the block.

The hole's location was center drilled, and a short pilot hole was drilled half way through the block using a 3/8" jobber's length drill. A slightly larger pilot hole was drilled completely through the block using a long parabolic drill. The block was then moved to the surface plate where the front and rear exit holes were verified to be within a thousandth of their 1.407" targets above the pan surface.

Upon returning to the mill, a long .604" drill was used to enlarge the hole which was then finished with a .625" reamer. Rechecking the finished hole's location revealed a rear-to-front downward-sloping error of .011" which was twice what I'd been hoping for. A minor tooth change in the timing gear set can compensate for the error though. Remarkably, the side-to-side error was only .002", and its average was used to scribe a centerline across the webs on the bottom of the block to later locate the crankshaft.

Next, the block was returned to the mill for the drilling and reaming of the eight lifter bores on the starboard side. My documentation shows these holes located immediately adjacent to the unfinished inner walls of the banks. Even though no dimensions were provided to locate them, the cylinder bores and head bolts were referenced to them. My lifters were positioned for a wall clearance of .015" and the dimensions in the documentation used to locate the cylinder bores and head bolt locations with respect to them. I selected a .995" separation between the cylinders and lifter bores which was an average of three different values found in my documentation. My dowel pin holes were used to reference the actual deck drilling operations.

The actual .251" lifter bores on the starboard side were drilled and reamed with no issue although the rear-most starboard bore ended up extremely close to a boss cast into the lifter valley. The purpose of this boss (and a few others in the lifter valley) aren't yet understood as they're not shown on any of my drawings. Hopefully their purpose will become clear later.

When it came time to rough in the 1.123" bores for the cylinder liners, I was concerned about the big discrepancy between the locations of my roughed-in bores and the existing cast-in holes. The 1" end mill that I was plunging wasn't even large enough to clean up the outside edges of the cast-in holes. The drawings picture the bores approximately centered on the decks, but a .995" bore separation wouldn't allow this in my block casting. I ran into a lot of hard material at the bottoms of the bores that affected their surfaces. The bores will be finished later in a different setup.

When I set up for boring the starboard cylinders I didn't notice the block was mounted backwards in the vise until I'd cut the first bore 1/16" deep in the wrong location. It turned out that the error will end up entirely within the counterbore by .002" when finished. I hardly ever wind up on the winning side of these mistakes.

The holes for the 8-32 head bolts were limited to a depth of .250" to avoid penetrating the water jacket. The same features on the port-side deck were machined similarly.

Some may be critical (and rightfully so) of the 'machining as I go' process I'm using without a fully modeled engine. I can only say that the modeling is so frustrating that I'm likely to burn out if I don't produce some tangibles along the way. - Terry

 

[GreenTwin]

This is an impressive display of coordinating multiple complex variables across drawings and castings.
Like tiptoeing through a minefield.
I have pondered building a V-8, but as this thread indicates so well, this is a very complex task even under ideal conditions with good drawings and good castings.
Hats off to anyone attempting this level of work.
Watching in awe and learning.

Edit:
I still say it would be easier to rework the engine in a correct 3D model, 3D print new patterns, and make new castings, instead of attempting to build from drawings and castings that are perhaps not as accurate as they could be.
I understand that the castings and drawings are in-hand though, and so the challenge is to make it work; so that makes for a REALLY interesting build, for sure.
Lets hope some sort of workable 2D drawings can be derived from this build that would working with existing castings for this engine.

.

 

[PTSchram]

We are in the company of greatness!

While my interests do not include model V8 engines, I am watching with interest as I have yet to fire up my KwikWay automatic boring machine. There are many similarities in functions and I have much to learn and watching this build will certainly help!