Handle blind nut

The last part of the day is the blind nut that holds the handle onto the lever. I would much prefer to purchase this part, but the closest equivalent I have found thus far is an internal Allen drive nut from McMaster, but because they are esoteric, they are expensive. However, the machining operations requiring different setups don't have to be done sequentially, so knocking out a bunch of these would be pretty quick.

I couldn't resist putting the three parts together.

Lever handle

The retainer plate took a long time to setup - nearly four days, most of which was getting up to speed with the cam software. Now that it is done, I decided to do something easy and rewarding.

Compared to brass alloys, of which there are quite a few, the array of engineering plastics out there is eye-watering. I presume that the lever, portafilter and valve handles are injection molded thermoplastic, but what kind I do not know. As designed, the valve handles cannot be turned on a lathe. The other two can be cut from stock quite efficiently however. My intention is to use black Acetal resin for the lever handle, but I only have white on hand - and they ain't gonna be white. I did have a rather nice piece of mahogany which has been waiting for a purpose for quite some time. It wasn't quite big enough in one dimension, so I split in two and laminated it with some birch plywood. 

You're not supposed to cut wood on a metal lathe because the dust quickly contaminates the oil film on the moving parts, but with a cover over the ways and a vacuum nozzle close to the cutter, I can contain the dust almost completely.

Profiling complete.

After the cutoff operation, the work piece is now unsupported at one end, so enlarging the bore is done with tiny cuts.

With a little sanding, some tung oil and about an hour and a half from start to finish including cam programming its all done! The new handle next to the original.

Retainer plate part 2

The foam test completed, I can move on to cutting the actual part.

With a small production run in mind, I made a jig to hold the stock which is then prepared with six holes, two for alignment with 1/4" dowel pins so the stock can be removed and replaced if necessary, and four threaded holes for the M8 screws that are required in the finished part and will do double-duty holding the stock to the jig.

The part after the program has run. The jig is intended to be semi-ablative - so the tool can cut past the part into the jig and I can skip the step of flipping it over and removing the base that would normally be held directly in the vice. However, there is another small offset error that left 2mm on the bottom of the part that had to be removed. 

A little clean up with a file and we are done.  :D  There are still a few small refinements to be made in the cam programming, but next time I should get it all right! For now, this one is close enough.

Retainer plate part 1

The project has moved forward a little over the last week.
I have finished up a few more parts for the group.

The first is what I call the retainer plate. Its function is to hold the group lever in alignment with the axis of the machine and to provide a stopper for the lever in the resting position. Originally, it is a casting with some secondary machining, but the material seems to vary - the two examples that I have may be either nickel or brass. Given that the function of this part is purely mechanical, the only reason I can see not to substitute aluminum is the possible differing coefficients of thermal expansion. For now, at least to check my design, I will be using aluminum and, rather than paying for another mold just yet, I am machining the part from solid stock.

My other preoccupation this week has been test driving some new cam software. So, with the twofold intention of protecting my mill and not scrapping a nice chunk of aluminum, I ran the program first on some high density foam. No worry about destroying expensive tooling if there is a collision with this stuff!

The problem with making parts like this is material removal. The final part is less than a quarter of the volume of the stock which means rather a lot of chips are going to have to be made in this process. Although it a little unorthodox, I am using an indexable carbide face mill as a clearing tool. At 1600rpm, the spindle on my machine just doesn't go anywhere near fast enough to make small diameter tools efficient. Here, the first rouging operation is mostly complete and the form of the part has started to emerge from the stock.

...moving onto a 1/2" corncob roughing end mill to remove the internal pocket and cut the external profile.

...and finishing up with a 1/4" ballnose to put the fine detail on the raised shoulder that serves as the lever rest.

The finished part in foam. I caught a number of mistakes with this test, among which was an offset problem with the cam program (note the four threaded holes are not centered correctly) and, more importantly, a few small errors in the part design. Prototypes are important!

CNC BSPP threading

To BSPP or not to BSPT

British Standard Pipe Parallel and British Standard Pipe Tapered.

ISO inch unit based standard with 55 deg thread flank angle. How an inch based thread ended up as an ISO standard I do not know...

The sizes concerned for the Aurora are:

1/4 - 19 (dash 04)
3/8 - 19 (dash 06)
1/2 - 14 (dash 08)
1"   - 11 (dash 16)

possibly also:
1/8 - 28 






For externally threaded parts, the major diameter is nominal size plus 1/4", thus 1/4 - 19 is 1/2" diameter (and 19 tpi).

Full discussion of how to identify threads and for the "dash" naming convention are here and here.

Discussion of carbide inserts for cutting BSPP threads

BSPP and BSPT threads are 55 degrees, not 60.

A good visual reference guide to threading standards.

A good discussion on Practical Machinist of how to cut BSPT and BSP - which is the same as Whitworth. 

Inserts:

If you select the Whitworth or BSP profile Carmex inserts, they will cut the full profile. Only the "partial profile" inserts leave the crests as-is.

So, Carmex "11 IR 11 W" or "16 IR 11 W" [for BSP, BSF, BSP, BSP] or "11 IR 11 BSPT" or "16 IR 11 BSPT" [for BSPT] or the equivalent Vardex or other competing inserts should do what you want.

Supplier (at $31USD each!) - (possibly not the best grade of carbide/coating for non-ferrous metal).
https://www.grainger.com/product/CARMEX-Threading-Insert-4PRZ6

Notes - a full profile insert is specific to the pitch and will cut radiused roots and crests. A partial insert will cut a range of pitches but leave flat roots and crests.

Carmex inserts seem to have the same naming convention as the Iscar, so the 11 IR 11 W example translates to:

11- insert size 11 (which is 1/4" inscribed circle)
I - Internal
R - Right hand
11 - 11 tpi
W - whitworth

Also, there seems to be some debate as to whether the radius roots and crests are necessary ("only important to amateurs"). My guess is that a sharp root will decrease the strength of the part, but that is highly unlikely to be an issue at the low working pressure in this application.


Possible inserts:

Explanation of ISCAR threading insert codes.
Good explanation of threading concepts, geometry and terminology.

Partial profile inserts (i.e. one size for a range of thread pitches with non radiused crests and roots)

ISCAR - External 55° partial profile, laydown threading inserts.
16ER A 55 IC908 - part number 5902222 
TPI min 16 - TPI max 48 - which would exclude the 1/2 inch

16ER AG 55
External 55° partial profile, laydown threading inserts.
TPI min 8 - TPI max 48

any of these carbide grades would do:

IC 508 grade
IC 228 TiN PVD
IC 908

Full profile inserts:

11ER 19 W   for cutting 1/4 - 19 and 3/8 - 19
11ER 14 W   for cutting 1/2 - 14

holder for all of these(?)


SER 1616 H16

Presuming that one holder can accommodate different sizes of insert (the 11,16 and 22 refer to the length of one side of the triangular inserts which inscribe a circle of 1/4", 3/8" and 1/2" diameter respectively).

References for CNC threading 

excellent explanation of code for CNC threading inch parts - however, while the code is specific to the HAAS or whatever he is using, the concepts are the same for the canned cycles in Mach3.

https://m.youtube.com/watch?v=e03pTbEBuGg

And now in Metric...

https://m.youtube.com/watch?v=e03pTbEBuGg


NYCNC guy's guide to thread milling with HSMworks and a Tormach

https://www.youtube.com/watch?v=a43S2y7Ccy8


HSMworks importing form tools
https://forums.autodesk.com/t5/support/hsm-2017-form-tool-does-not-show-on-import/td-p/6311059


G76 canned cycle explained
http://www.helmancnc.com/mach3-turn-g76-threading-canned-cycle/

Lever part 2

Here are a couple more images of the process of machining the lever, which I managed to complete last night.

I attempted to limit the machining processes as much as possible to left hand cutting operations to keep tool changes to a minimum, but the inside radius of the handle needs another tool and is best cut in the other direction.

Here is the scratch pass of the threading operation. I paused after this to check that the thread pitch is indeed 1.5mm.

Then, some 47 passes later, the threading is pretty much done.

A fair amount of sanding and polishing later and here is the finished part on the right with the original and the steel from whence it came on the left. As a side note, the threading at the small end of the lever that keeps the handle in place turns out to be imperial 5/16 - 18. Probably because they could't get the rather exotic blind nuts from a metric supplier. This makes it the one single exception to the otherwise absolute metric rule of the machine.

Group lever part I on the new CNC lathe

Over the last few weeks, while I have been waiting for the foundry to setup and make the first casting, I have been converting a lathe to CNC in order to fabricate some parts in house. This seemed sensible for a number of reasons among which is prototyping, which is expensive to have done, but possibly also for some small-volume production. The plumbing parts for the boiler are good candidates as they are expensive and in a few cases impossible to find. Additionally, the ones that I can purchase from my local supplier (and from what I have seen online) have no material certification - which means that they may or may not contain lead. 

So yesterday, after close to a month of modifications and wrangling new software I started cutting my first part: the lever handle.

This is just a piece of hot-rolled steel that I had lying around. I will be using cold-finished 1018 steel or I may switch to 7075 aluminum if I can find it cheaply enough here in Canada. 

The first couple of passes taking off the nasty foundry scale before preparing one end of the stock.




Roughing passes on the straight and conical parts of the handle.

Today's goal is to try to finish the profiling and threading.

First foundry casting

The first casting from the foundry! The square block at the top is the riser which provides a reservoir of molten metal that is drawn into the casting as it solidifies and shrinks. There seems to be a few slight dents in the cast, likely the wax pattern was dinged before it was put into the investment. They should come out in the polishing, though are couple are pretty deep. Now it is off to for machining.

This leads me to a discussion of the material. This is a functional prototype made from a brass alloy similar to C69400, which is not an ideal material for a couple of reasons. The first is that the lead content likely exceeds allowable levels (0.3% in the US). The second is that C69400 has no additives to prevent (or rather retard) dezincification, the process whereby the zinc is stripped from the metal leaving a weak and brittle copper sponge which can ultimately lead to failure of a part at pressure. Dezincification resistance is especially important for espresso machines because of the operating temperature, the acidity of coffee and often, due to the use of water softeners, the mild salinity of the water itself. 

From the excellent copper.org website:

[i]Sand-cast faucets and other plumbing components have traditionally been made from leaded red, semi-red and yellow brasses. The most common plumbing brass, C84400 (also known as 81 Metal or 81-3-7-9) contains nominally 7% lead. The most popular red brass, C83600 (85 Metal, 85-5-5-5), contains nominally 5% lead. Permanent mold and pressure die castings of plumbing components are also commonly made of the leaded yellow brass alloy C85800, which contains nominally 1.5% lead.[/i]

Lead is added to brass to improve machinability. It acts as a lubricant and causes the chips to break into small pieces while it is being cut. Worse still, because of the way the lead crystals form as the liquid metal solidifies in the mold, the concentration of lead is highest at the inside surface - i.e. where it comes into contact with the water. The unfortunate conclusion is that it is highly likely that both my machines (and indeed all vintage espresso machines), help me meet my recommended daily dose of lead in the morning. This was just the way things were was until California passed its law in 2006. Since then, considerable effort has been made to find alternatives to leaded brass.