Field Update — September 2026: The Foundry Equipment Arrives, and a Simpler Way to Spec Your Block

This is the first of what will be a monthly field update. The idea is simple: once a month we tell you where the Gen III HEMI CGI block program actually stands — what moved, what did not, and what we learned. If you are waiting on one of these blocks, you should not have to guess.

We will not dress it up. Some months the news will be a machine landing on the floor. Some months it will be a supplier lead time going the wrong way. Both are worth telling you about.

Where we are on equipment

The block design is done. The thing standing between a finished design and a block on your engine stand is the shop that pours and machines it — and this month, that shop started arriving.

The SinterCast process control equipment is here. It shipped from Katrineholm, Sweden, cleared customs, and is on our floor. We are now building the production sampling rig around it — the fixturing and workflow that let an operator pull a melt sample, get an answer, and act on it without slowing the pour down. Of everything arriving this year, this is the piece that decides whether the rest of it matters, and the next section explains why.

The 2006 OKK HM600 horizontal machining center is bought and paid for. Install and commissioning are on the calendar with Merrifield Machinery Solutions, and that scope deliberately includes laser leveling and a ball bar geometry check with a written report — more on why below. Horizontal is the right architecture for a block: chips fall away from the cut instead of packing into it, and a pallet lets us reach multiple faces without breaking setup. Both matter when the part is a heavy iron casting whose bores, decks and lifter bores all have to agree with each other.

The melt furnace lands Monday, September 14, and a 150-ton, 3.2-metre press brake comes in with it. The furnace is the obvious one. The press brake is not — and it is there for a reason we will get to in the fixturing section.

Where that leaves us: by the middle of the month we will have melt, treatment and measurement under one roof, and the machining side has a commissioning date. We are not going to give you a first-pour date we cannot stand behind — when those three have run together on our own floor, you will read it here. If it slips, you will read that here too.

Why the casting equipment is the hard part

It is worth explaining why we are being deliberate here, because it is the least obvious part of the program.

Compacted graphite iron is not simply a stronger grey iron you can pour with the same setup. What makes CGI what it is — graphite that forms in worm-like compacted flakes rather than grey iron's sharp flakes or ductile iron's spheres — depends on holding the melt chemistry inside a narrow window. Push the treatment too far and you get ductile iron. Not far enough, and you get grey iron. Neither is what we designed the block around.

That window is narrow enough that you cannot reliably hit it by recipe alone. It takes measurement and correction during the pour itself. So the equipment decision is not really "which furnace" — it is which process control we can run, how repeatable it is casting after casting, and whether we can prove it on every serialized block rather than on a sample.

That is what the SinterCast unit is for. It pulls a sample of the melt and runs a thermal analysis on it — reading the cooling curve to work out where that melt actually sits — and then tells the operator how much magnesium and inoculant to add before the iron goes into the mold. You are not pouring and hoping. You are measuring, correcting, and then pouring, every single time. It also logs what it did, which is what makes a certificate mean something.

That is also why the sampling rig is real work and not an afterthought. The measurement is only useful if the operator can take it fast enough that the iron has not moved on by the time the answer comes back. Building that rig properly now is what turns a good instrument into a repeatable process.

This is also why we are doing it ourselves instead of sending the pattern to a contract foundry and hoping. If the material properties are the entire argument for the product, we want our hands on the process that produces them.

A block that measures 450 MPa minimum tensile strength on the certificate and 450 MPa in your engine is the whole point. Getting there is an equipment and process problem before it is a casting problem.

Buying a twenty-year-old machine on purpose

A 2006 machining center is a deliberate choice, and it is worth saying why, because it runs against instinct.

A machine tool's accuracy is not a number stamped on it at the factory that quietly decays on a schedule. It is a geometry: squareness between the axes, straightness of travel, backlash in the drives, and how level and stable the whole machine sits on the floor. Nearly all of that is measurable, and most of it is correctable. What you are really buying in a well-built machine is mass, rigidity and good ways — and those do not expire.

So before this one cuts a block, it gets qualified. First we laser level it. A machine that is not sitting true twists its own castings, and every cut it makes inherits that twist. Then we run a ball bar test: the machine traces a circle while a precision telescoping bar measures the path it actually took against the perfect circle it was asked for. The deviations are diagnostic rather than just a pass/fail number — backlash shows up as steps where an axis reverses direction, a squareness error shows up as an oval, and running the circle in both directions separates errors that look alike from one another.

The goal is to qualify this machine against our global machining tolerance before it ever touches a customer's block, and then to re-run the same test later and show it still holds. A used machine that has been leveled, ball-barred and documented is a known quantity. Any machine that has not been is an assumption — and assumptions are not what you want holding your deck flatness.

Tooling and fixturing: the part nobody photographs

Buying the machine is the easy half. The harder half is everything that holds the block while the machine cuts it.

We are building OP 10 fixturing right now — the first operation, where a raw casting with no machined surfaces on it gets located, clamped and given its first datums. Every operation after that references what OP 10 establishes. Get it right and the rest of the sequence has something honest to work from. Get it wrong and you spend the next five operations chasing an error you built in at the start. On a 250-pound iron casting that has to hold bore-to-deck and lifter bore relationships across its whole length, this is the operation that sets the ceiling on everything else.

Production tooling is 95% designed, with the tooling kickoff scheduled for the week of September 14. That is the point where the design stops being a model and becomes cutters, holders and a proven cycle.

This is also where the press brake earns its place. Fixture bodies, guarding, chip pans, tooling carts, the sheet metal work around a machine cell — every shop needs a steady supply of it, and every shop either makes it or waits on someone else to. We would rather make it. The same instinct that says pour your own iron says bend your own steel: if you can build the thing that holds the part, you are not waiting on a vendor's lead time to fix a fixture that is not working.

None of this makes for a good photo. It is also the difference between a shop that can machine a block and a shop that can machine the two hundredth block the same as the first.

Four blocks, four pages, no configuration required

The other thing that changed this month is on our side of the screen, and it is a fix to something that was making your life harder than it needed to be.

Until now, every block — all 176 valid combinations of deck height, lifter package, timing drive, cam tunnel, bore, and finish — lived behind one page and three dropdowns. If you already knew exactly what you wanted, that was fine. If you did not, it was a wall.

So the four configurations we get asked for most now have their own pages. One build, one price, nothing to configure:

• ACE STREET — $4,200. The drop-in. Standard 9.285" deck, stock lifters, factory VCT and stock cam retained, 4.090" rough bore. Fits 5.7L and 6.1L. Nothing else in the engine has to change.
• ACE BOOST — $5,000. Tall 10.75" deck and 4.210" bore for forced induction, with factory VCT kept so you hold onto stock-style timing control under boost.
• ACE DRAG PAK — $5,600. Drag and bracket spec. Standard deck, 937 roller-guided splay lifter bushings, non-VCT front drive with a 60mm roller cam tunnel for big lift.
• ACE MAMMOTH — $6,700. Our largest bore at 4.300", plateau-honed and ready for ring fit, on a tall deck with 937 splay lifters and the 60mm roller. Naturally aspirated only — see below.

The full configurator is still there for everything else. It just is not the first thing in your way any more.

Two things worth knowing before you spec one

Bore size does not change the price. A block is $4,200 plus what you choose for deck and lifter package, plus what you choose for timing drive and cam tunnel, plus $300 if you want it plateau-honed instead of rough. Bore is free. Take the bore your build actually wants, not the one you think you can afford.

4.300" is naturally aspirated only. At our largest bore the cylinder walls are specified for NA use and are not rated for a supercharger or turbo. If you are building boost, take 4.260" or smaller. We would rather tell you that here than have you find out on the dyno.

On rough versus honed: a rough bore ships .003" undersize so your machine shop finish-hones it to final size and can match the hone to your ring package. Plateau-honed is finished here on the same CMM-verified equipment that cut the block, ready for ring fit. Honed costs $300 more and usually saves more than that in shop time. Neither is the right answer for everyone.

What is next

Next month's update leads with four things: the furnace in and commissioned, the production sampling rig finished and running against the SinterCast, the results of leveling and ball-barring the OKK, and where the tooling kickoff landed. If the first pours on our own floor go the way we want, you will read about them here first. If they do not, you will read about that too. That is the deal with this series.

Every block we ship will be serialized and go out with its own CMM report and full QC documentation covering both the casting and the machining. Building the capability to do that consistently is the work in front of us, and it is most of what the equipment arriving this month is for.

If you have a build in mind and are not sure which configuration fits it, call us before you order. You will reach someone who knows the engine. That is worth more to both of us than a returned block.

Eric Cunningham
ACE INC · Brighton, Michigan
(615) 924-2737 · eric@aceinc.shop