The Kneeslider

Doers Builders and Positive People

Aluminum Block to Engine Block

By

Aluminum block before machining
Aluminum block before machining

Here’s how Kirkham Motorsports, the guys that build those aluminum bodied Cobra roadsters, put a big block of aluminum on a diet. You start the workout with 30 hours of CAD work followed by 50 hours of programming and finish with 30 hours of CNC. This exercise formula trims the overweight 386 pound block of aluminum down to a svelte 64 pounds of engine block.

Yes, you’ve seen CNC machines at work before, but this isn’t just a demo, they’re building a real engine and if you like engines and tools like most of us around here, it’s worth a look.

Video below:

Posted on Filed Under: Engines, Workshop & Tools

-- Subscribe to The Kneeslider --

Be the first to know when something’s new!


Be sure to check out The Kneeslider’s motorcycle manuals and literature listings.

Your purchases through ebay links on The Kneeslider may earn a commission for this site.



Comments

  2. Impressive!

    I am actually looking for infos on how to build one-off engines. What material to use, what heat treatment needed, what tolerance or surface finish for what etc etc…

    Does anyone have any info on that? Or a book maybe? Looked in amazon but nothing came up.

    Thanks,

    Marc

  4. Probably sell it back to the supplier for remelt.

    CNC = magic. What a way to make a motor! Just apply tons of talent, off-the-shelf CAD and CNC software, set up a high-speed mill and a few other machine tools, and you’re making whichever motor you want. I wonder if these guys would run off a nice BSA A65?

    • A65? omg they wont go away will they. I had one of these bad boys in 75. Twas a A65L ’72, you know the ones being sold off ‘cheap’ (£425) with tiny orange/white tank. I set off from Dorchester for the MCN bike show in Dec 75. Half way up M3 red oil light came on with engine knocking. Pulled over to consider things. Decided to continue with journey rather than give up. If it blows up – tough. Amazingly it got to the show and back home again – I think I was more bust than the bike. Turns out the infamous crank shaft sludge trap got blocked up cutting off oil to big ends (thats a really really bad thing to happen). Big end shell locating lips tore off after shells seized on crank pins, forming new ‘wide tolerance bearings’ in the conrods !!

      Anyway, IMO, I think the series 2 BSA Rocket 3 would offer much greater rewards for the effort if undertaking such a project of making new engines using CNC. I often wonder what an old type BSA or Triumph motor would be like if made on a modern CNC – all the incredible tolerances achievable, modern materials and quality assurance etc. Instead we have a bunch of mega rich bankers being paid billions for punching numbers into spreadsheets. Fantastic.

  6. Great stuff to do a one-off project or to build a new part for a long defunct classic machine. Especially where money is plentiful.

    Practicality would dictate a reversion to the more classic method of casting the rough block and using the CNC to do the finish machining of the critical surfaces. The machined exterior may certainly look pretty, but does it justify the additional expense? As a show piece, yes. But buried under a hood, I doubt it.

    At the same time, it’s a great work of art.

    • The reason for machining from billet is simply strength and resilience. When forming alloy, high pressure die casting is 4 times stronger than sand cast. Machining from billet is 10 times stronger. As a result, you can cut down heavy members weight by a considerable margin and still have more strength. When Harley (No, I am not a Harley fan. I ride a Goldwing) went to die casting on the Evolution engine they did not decrease the thickness of the material from sand cast. As a result, the Evo proved to be one of the best engines ever made and it took a lot of abuse in comparison to a shovel or pan. By extension, people like S&S and Murch went to billet cases which is why we have the killer strokers available today. This block is probably 1/2 the weight that a die cast would be and a third what sand cast would be and is probably able to take 3-5 five times the load without failure. If you are talking serious racing or high end cars the cost is justified by using the best.

  7. This is extremely impressive!

    @Hawk:

    Casting starts to make sense if you’re making more than one. To make something like an engine with the sand casting process you would need to make a wooden buck, core boxes, etc. Essentially, you would need to mill a lot of stuff out of hard maple. Milling maple is a little faster than aluminum, but not by a massive amount. Then when your part is cast, you need to machine the bores, the mains, the mating faces, etc. If you’re only making one, milling it from a billet makes more economic sense than casting one.

    @B50 Jim:

    You’re completely right about selling the chips and scrap back to the supplier or scrap metal buyer.

    The setup you’re talking about is pretty spendy. In software alone, you’re probably looking at $15,000. Factor in another 2-3 grand for a computer with enough processing power and memory to generate your tool paths. I can only speculate about that machine, but I would estimate it to be in the hundreds of thousands, if not above a million (all figures stated are for new equipment).

    • @ BAMF: Patterns don’t need to be that hard if the individual packing the green-sand uses reasonable care. If serious production volumes are going to be run then you’ll want to “cast” an aluminum pattern anyway. I’ve made patterns out of pine, MDF and even foam/Scotch tape. The softer materials take shape with less effort.

      Anything “billet” looks expensive and “track-proven”…but to parrot Todd, the flowing lines and relieved contours of a well developed casting “looks right” to me. Swiss-cheese anyone?

      • The one foundry I’ve worked with stated that they wanted maple for the patterns. We were making some very small parts, so that may have had something to do with it. They also ran a fairly automated system and were running a lot of different molds that were all roughly the same size.

  8. Marc look for two books titled Internal Combustion Engine in Theory and Practice: by Charles Fayette Taylor. There is a lot of good material in there.

  9. One thing not mentioned for far is the benefit of a forged block in certain applications. For example, Formula One will be going to a four cylinder turbo format in 2013. A forged block would seemingly be more dimensionally stable than a cast one at a given weight, and stronger than cast when used as a structural member of the chassis. Given the cost/effectiveness mind-set and low quantities involved in F-1, the CNC route (pun intended) might be less expensive. Just a thought.

    • Patrick, this starts out as a forged block (billet) but gets machined down into its form. Unlike traditional forging for a finish part machining this block cuts right through any grain that occurs from the forging process. A (non porous / inclusion) cast part will have a grain that flows around the shapes, adding strength. Casting also allows for more complex shapes, if required, without special or modified tooling or where it might be otherwise impossible in a machined part.

      Likely this was just done to show off the capability or a deep-pocket customer wanted it for his billet Cobra. There are new, cast aluminum 427 blocks readily available for less than $5,000.

      -todd

      • “A (non porous / inclusion) cast part will have a grain that flows around the shapes, adding strength. ”

        I’m sorry, I believe you have mistaken this with forging, maybe it was a typo? Depending on cooling rate, one should see non uniformity in the grains of cast parts, however, I do not believe these non uniformities show any great deal of strength increase (unlike seen in the forging process).

        • From what I’ve been led to understand about casting is that rapid crystalline growth (equiaxed) increases the strength of aluminum parts. In order to achieve equiaxed crystalline growth parts must have high surface area and thin cross sections. High surface area is achieved by wrapping the shape of the mold close to the desired finish size and adding features like ribs or grain refining additives. Making parts thin and adding ribs, greatly increases the strength of a part while also making it very light.

          Machining from a solid block will cut away the desirable equiaxed surface crystals exposing the columnar (large and directional) crystals that are usually a detriment to strength. Parts will need to be made thicker because of this (and because they might not be able to get a tool into an area to cut away metal) making the part heavier and not as strong.

          Die casting is a nice process that provides a light, strong and beautiful, organic shape that is difficult to replicate with machining. It does have its drawbacks, mainly development and equipment time and costs.

          -todd

      • The customer is Larry Ellison. The chassis & most of the car is also billet. An amazing project, its on Kirkham’s website.

  10. At one point they were talking about machining the “lifter valley” – does that mean that this high-tech engine runs push rods instead of OHCs?
    I didn’t see anything that would indicate space for a cam chain.

    • Yep, the Ford FE is a plain old pushrod V8 that dates to 1958. It was the engine theat was original to big-block Shelby Cobras of the mid ’60s.

  11. As a non-engineer, why would this not be a good idea for a limited production builder, say Aston Martin, Panoz, Koniggsegg, even SCC or perhaps a new American motorcycle company? I know that the initial investment is expensive, but for anyone who brews his own beer at home will tell you that this is the same thing, but over time of making 5 gallon batches at a time, the costs quickly come down per bottle.

    When the excess aluminum is sold back to the supplier, this means that an economic arrangement can be made that could be negotiated to favor both parties – an engine builder who wants to buy a certain weight of finished product (make enough engines to weigh 400lbs) and the aluminum block supplier has a steady customer.

    Any MBAs on here also see this a feasible in certain applications?

  12. I`m not an MBA but I do remember reading an article about TVR`s new-found ability to develop thier own power-trains for small production runs, thanks to in-house CNC capabilities. I think they`re gone now.

  13. Cool stuff, the top and ends of the engine look to be FE, but the bottom is not “Y” block style. The block looks to have 4 bolt mains and with out the “Y” skirt there will be no cross bolts. This made the block even lighter and easier to machine.

    The guy who figured all this out is pretty smart and knew what they wanted and did not just make a copy of someones work.

    But by the time you put the sleeves, studs and main caps on it, the block should weight a little more.

  14. random blather from an aerospace engineer at home nursing a cold:

    A lot of the structure I work on is CNC. In a production environment, it’s a lot less glamorous and magical than you might think. We don’t get perfect parts every time. Set-up error, cutter breakage/wear, material issues, and other bits of reality lead to a steady stream of parts that don’t match the engineering dataset, require rework, or are scrapped.

    Parts that require this much machining are avoided. Early in a production program, we’ll make parts this way as a stop gap. It’s flexible enough to allow us to resize parts if they aren’t sized correctly. It doesn’t require as much lead time as forgings which require dies, forging trials, and part qualification testing.

    I didn’t watch the video with the sound on, so, I’ll assume the comments that say this is forged block is correct.

    The increase strength from forgings is in proportion to the thickness. Dies whacking down on the material stretches it, aligning grain structures, forming layers within the material that give it strength. The strength increase is greatest in the directions parallel to those layers (called “long” – L and “long transverse” – LT). The strength through the thickness (“short transverse” – ST), perpendicular to those layers, is less than the L and LT directions. On our drawings for highly loaded parts we identify the grain direction on the part and restrict the thickness of the forged material (so they don’t take a thick block and split it in half).

    This engine block is a “hogout”. For something this thick (>6″) we would typically rough machine a non-forged block, heat treat, and then finish machine to final size. For this application, the exposure to elevated temperature from coolant, oil, and radiate/conducted heat might negate any forging/HT strength benefit.

    For a given metal alloy, the greatest strength typically comes from a precision die forging or a near net forging (that requires a little machining to finalize the part geometry). Forged block is fine if the part isn’t very deep and the loads are in line with grain direction.

    I work on primary load path structural (highly loaded) parts. Typically, we don’t use castings. Wrought materials typically have higher strength. With castings there’s the issue of processing. Every pour has a chance of errors occurring. When we buy material from a supplier, their quality control and testing that assures the strength properties. Occasionally something gets past them and we have to assess the impacts of this out of standard condition. With castings, after the pour there is testing to verify the properties and quality of those castings. If they find a problem it comes to us. “We have some porosity in this part of the casting. Can you buy it off or do we scrap the 20 parts that were part of this pour? Oh, and the factory says that scrapping these parts will cause a two week slide.”

    Britten cast his own motors. Yamaha has been doing great things with vacuum assisted casting. For one-off production, lost foam is pretty economical. The hack-o-sphere could provide laser cutting of foam sheets that can be stacked, glued, and ready to be dipped in investment. It can be cast DIY or sent to a casting house.

    Potentially you could make a lighter engine block by using something other than a casting. But, to take advantage of greater strength from another material form, you have to do the analysis to determine where you can take away material without sacrificing strength or reliability. If the overall geometry of Kirkham’s block isn’t different from the production cast aluminum blocks mentioned above, then there wasn’t any weight benefit. There might not be a strength benefit for an assembled engine if the weak link is some other component like crankshafts, conrods, bearing bolts, etc.

    Kirkham manufactured the engine this way because they had the tools to manufacture it this way. It’s an impressive display of machining and capability. It isn’t the best way to make an engine… just one way. If they had a more affluent client who wanted a forged engine block they could have spent a ton of money on making one-off dies for that purpose. They could have commissioned a titanium cast block that was HIP’d to make a much lighter, stronger, and avoid the thermal expansion and galvanic corrosion issues of aluminum (with the iron liners).

    There’s a kind of fetish for aerospace materials/processes (CNC aluminum, carbon fiber, titanium) in the general public. Working with those materials everyday, they lose some of the gloss. The machining pattern you see with “billet” parts are something we remove. That surface roughness reduces the fatigue strength and reliability of the part. For fatigue critical parts, those lovely patterns get erased by shot peening. CFRP is light and strong, but, fiberglass can be lighter and stronger. The downside of glass is that it’s less stiff, so, more is used to keep it from deflecting under load. In aerospace, the pretty fibers are buried under primer and paint because UV degrades the epoxy and the strength. Titanium is great stuff, but, it’s difficult to machine. When titanium is in a joint stack it wears out drill bits and often leads to out of tolerance holes when Ti shavings get caught by the bit and gouge the other parts in the stack.

    • Bob, you should work something out with Paul as a regular contributor. That was VERY GOOD reading – the kind that keeps me coming back for more.

      Thx.

      • Paulinator, I barely have time to comment on blogs, let alone contribute, but thanks for the compliment.

        Another comment about material, something I didn’t realize until I got to a production program and started to see rejection tags. All metals contain some level of residual stress from it’s processing history.

        This starts from when the metal solidifies from liquid. The outer surface solidifies first. The center is still molten and will contract as it cools. The causes residual tension in the center, residual compression in the surface. It can even cause voids/porosity in the center of the block when the “vacuum” created by contraction pulls dissolved gases out of solution from the alloy forming bubbles. These residual stresses can be mitigated by stress relieving processes, but, they’re still present at a reduced level. Other processes like forging, stretching, extruding, and heat treated introduce residual stresses. Some of those stresses are responsible for the increased strength from that processing.

        Machinists know from experience that those residual stresses cause spring back and warpage. Machine one surface from a thick heavily processed plate/block and the loss of that surface’s residual stress can cause the plate/block to potato-chip when it’s unbolted from the table. It’s worse when the machined detail is unsymmetric when most of the material is close to one surface. Machining specs assume a certain amount of allowable warpage and allow a certain amount of “straightening” (bending the part back into shape). Most structural parts get some stress relief process after machining.

        Forged blocks are unique. For die forgings, the parts have a defined geometry. There will be specific dies and numbers of die strikes to whack the part out of that starting piece of metal. Typically, the final die strike for the part will be on one matched set of dies. For forged blocks, the geometry isn’t defined. Any given press can deliver a specific impact force. Bigger presses not surprisingly are more expensive than smaller presses. Because the forged block can have a large surface areas it isn’t always possible to strike the whole block at one time. Often the press uses bars to focus the impact on a band of material on the block. This is similar to a blacksmith hammering or using a drop hammer forge on a piece of metal. Individual die strikes overlap, not always uniformly. We tend to see more post machining warpage with forged blocks. Also, the benefit from the forging process varies through the thickness of the part. Close to the surface, there’s lots of material displacement that aligns the grains and forms layers. On thick blocks there’s much less material displacement in the center and it isn’t as strong.

    • Bob ! YOUR THE MAN !!!!!!!!!!!!!!!!!!!!!!!!!
      I am awaiting delevery of a Optimum 46 milling machine that I am converting to CNC. I learned so much from you reading so little.

  15. Back in the old days we simply got some plate stock and bar stock and cylindrical stock and welded it all together to manufacture our fine Crossly engine blocks.
    So let that be a lesson to you young wippesnappers!

  17. I don’t know jack squat about engineering.. or engine building really (learned a little from reading all the comments just now). But I found this video to be quite amazing, mainly because of the effort that went into it all… Its an amazing (not really in the magical sense… ) process.

  18. CNC isn’t magic — a whole lot of work goes into writing a program. Things do go wrong and there can be some spectacular crashes. Those safety doors are there for a reason. Still, it can do things a human machinist can only dream of, and it never gets tired. If the program is right, it turns out part after identical part. Still, for making any quantities of engine blocks, casting is all-around better.

  19. @Tom

    SSC does use a billet block. They were one of the first companies to do it. The only company that I have known to do it before them was Dart. Aston makes too many cars, Panoz uses Chevy engines I believe. It would be good for Koenigsegg but they already have the tooling for cast.

    • Aston makes too many cars? Economies of scale would make doing so more cost effective. Its when you make one car a year (like SSC seems to do) that the cost becomes astronomical. Kirkham responded that chips can be recycled for recycling and creating a new ingot. This being the case, economies of scale could make this process more common than it is today.

  20. Beautiful.
    I could spend a whole day in your shop and look at that work of art.
    Thank you for showing this.
    Jörg.