Steering Knuckles

Last semester I had an opportunity to help out the SAE Formula racing team at my school by making some parts for them. To set the scene, every year the team designs and builds a small formula style car from the ground up. This requires an inordinate amount of effort from the team members that culminates in competitions over the summer. For various reasons, I never joined the team but I often work in proximity to them because of my work managing the various CNC machines in my lab on campus.

Finished front knuckle. Large hole diameter is ~3.5 inch

Last spring, in the middle of crunch time for the team making parts, our primary 3-axis mill (Hardinge V1000) had an issue with the tool changer. We don't know what cause the issue but we suspect a crash or similar incident that wasn't reported. It ended up with the ATC timing being off and the machine needing some new parts. While the parts were being shipped, the team had to scramble to get all their parts done on the other available machines. 

In addition to implementing some new policies as a result, we decided to help them team out in our lab and offered to make their steering knuckles (uprights) on our 5-axis. This is the same Okuma 5-axis on loan to us I have referenced in some of my previous posts. Because of the cost of the machine, and the fact we don't own it, the owners requested that I be the only one to use the machine while on loan to us.

Finished rear knuckle. Large hole diameter is ~3.5 inch

Long story short, I offered to make the team's steering knuckles on the 5-axis to save them some time. I made these before the turbines I previously wrote about, and these parts were my first real parts to come off this machine. In total I made 8 parts (2 each front and rear, left and right), with 2 proving parts, 1 each for front and rear. 

The parts started out as big chunks of 7075 aluminum. The blanks were prepped with a simple dovetail for holding in the vise, no other prep needed. Roughing and finishing of more than half the knuckle is possible with the first setup. All critical features and dimensions are completed in the first op, except for the back bearing bore. I had some real fun here with a Sandvik 590 face mill, blasting off extra material like it was nothing. 93 cuin/min and the machine barely noticed, with sustained spindle loads maxing out at 80%. The only down side was spraying all those big chips around the enclosure scratched the plastic window on the door.

Stock before op 1

Stock after op 1

I completed the first op on all parts before second op, mostly because of the fixture I decided to use. The team had shared some pictures and description of how they made these parts in the past on a 3-axis. The basic idea was square up all sides of the stock first, then rough out the top and bottom thirds of the part in 2 ops. With the bores finished, and some other holes established, they could clamp the part through the bore, using some extra holes for rotational alignment, and finish the parts. I opted not to use this method because it required the stock to be squared first, taking a long time. Their method also used extra holes for rotational alignment, which I didn't really trust (I'm an indicator man). The fixture I used ended up saving a lot of time and effort. 

The team's original method of fixturing

Having roughed out 90% of the part from the first op, I flipped the parts over and secured them using an expanding bore clamp. These are fantastic fixturing tools that can get you out of tight spots. Not only do they not require any external clamps, gripping only from the inside a hole or pocket, they also provide full support behind thin walls. They are also good for situations like mine, where you want to clamp in a blind hole. Because of my vise, I couldn't make the full size bearing hole through the whole part. If I did, the part would have fallen off the vise. Instead I drilled a small through hole, only enough to fit in the hex wrench to tighten the bore clamp. The bore clamp located the part in every axis but C, which I indicated in by rotating the C axis of the machine, and setting my work offset. 

My method of fixturing

This worked great. I went a little easier roughing on the second op just to be sure, but there was nothing to worry about. The clamp held strong and didn't rotate. I didn't measure concentricity of the two bearing bores (because there was no tolerance on the drawing), but I skim cut the bore clamp in place, so it should be very good. This clamp worked for all 8 parts, no needing to swap fixtures. It also gives

My method of fixturing with part

Off the machine the parts needed very little finishing, only a small amount of debur. There were some tapped holes for wheel speed sensors that I had drilled, but not tapped in the machine.

All the fixture design and programming was done in Fusion360. Fusion is great for positional 5-axis work, which all most all of these features were. For my thoughts on Fusion's ability for full 5-axis, read my turbine article. One thing that really came in hand with Fusion is the ability to write your own post. I modified the stock Okuma 5-axis post the comes with Fusion to make it work much smoother with our machine. Little extra safety moves here and there, custom tool change cycles added, or even adding multiple tool length offset support I could do fairly easy, and I don't even know Java script the post is written in.

My time to program, set up, and run these 8 parts, including modifying the post processor, was ~85h. This is down from the 200h the team had documented previously making only 4 of theses. Obviously having the 5-axis and a 15k spindle really helped with this. But equally using the right tooling, feeds and speeds, and clever fixturing saved time. This job was a good experience for learning 5-axis programming and operation.

As a thanks, the team put my logo on their car as a sponsor.

8 parts and 2 proving parts on the far left

An Engineer's Spoon

This project came about from a need for a kitchen spoon. Now you would think you could just buy one of those, but where is the fun in that? The bowl is machined 304 stainless steel, with an Osage Orange handle. I was inspired by a limited edition Le Creuset cast iron skillet I came across a few years ago. It was ridiculously expensive, but had a nice aesthetic where the wooden handle blended smoothly with the metal body. Around a year ago I needed a nice kitchen spoon. Growing up with wooden spoons, I gravitated towards that as a choice, but there weren't really any I liked and over time the wood really wears because its never hard enough.

The alternative is stainless but the handles on those aren't comfortable. Remembering the aesthetic of the skillet I liked, I designed a spoon. That spoon was never made. I designed it more as an experiment and didn't really think about how I use spoons when cooking and what manufacturing methods I have available to me. I shelved it and forgot about it for a bit. 

Earlier this year, we finally got our Okuma 5-axis up and running (same machine in the impeller post) and after a busy semester learning a new machine, control, writing a post processor, and making parts (more on those parts to come), I had time on the machine to make a spoon. I dusted off the old CAD model and made a bunch of changes. The bowl was slimmed down to fit the stock I had, the leading edge was thinned out to make scraping the bottom of pots easier, and some of the lines and tangencies were refined. I did all the modeling and programming in Fusion360, and since the first spoon design, they added some more features that helped me refine the look. 

Programming this was a good learning experience for 5 axis work. Stability was a big issue throughout as well as tool access. Off the machine I was quite happy with the finishes. This is due to the superior quality of Okuma, as well as having some really top notch Sandvik tooling. After machining, I just sawed the part off the remaining stock. At this point an interesting issue was presented to me; how do I drill the hole for the handle when there are no reference surfaces, flat or parallel edges or really any way to hold this part? I created an aluminum tool that was machined to be a perfect negative of the bowl. This gave me support over a large area, as well as rotational alignment to clamp it in tall jaws seen below. From there it was easy to indicate the round shank of the spoon and drill the hole. Last was the handle, which I completed over the summer. Very straightforward compared to the bowl. Osage Orange was used because it is very hard, strong, and moisture resistant. Ideal for use in a kitchen without the need for any finishing. Also it matches a kitchen knife handle I made a few years ago.

Having used this spoon for a few months now, I do have some lessons learned. First, weight distribution. The spoon has a heavy metal bowl and a very thin, wooden handle. Its front heavy. If I had thought of this, I could have checked center of gravity in CAD. Spoon still works fine but it takes some getting used to the unique weight distribution. Second, Loctite 380, my favorite adhesive, is not good at bonding to wood in high humidity environments. The joint has come loose over time, but the machined fit is so tight, that once the handle heats up at all in a pot, it swells enough that it won't come off; so no long term issues. Last, I would use a harder material next time. Maybe a 17-4 PH. Something that resists tapping on the edge of a skillet a bit better. 

Now all I need is a fork.....