Science desk
< September 9	<< Aug \| September \| Oct >>	September 11 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

September 10

Machining

Suppose I want to do the following operations (starting with a short hex bar of the appropriate length and cross-section): (1) Machine one end down to a cylinder, and the other end down to a truncated cone (leaving a hexagonal flange in between); (2) Hollow out the cylindrical end so that it matches the wall thickness of a Schedule 40 pipe; (3) Drill a smaller-diameter channel along the axis of the truncated cone such that it connects with the hollow cylinder made in steps 1 and 2; (4) Cut a male National pipe thread on the cylindrical end; (5) Cut a female National pipe thread inside the narrow channel made in step 3; (6) Thoroughly deburr all surfaces. My question is, what equipment would I need for these operations: (a) Lathe; (b) Drill press; (c) CNC milling machine; (d) Some combination of the above (please specify); (e) If this task also requires other machines or hand tools not listed above, please specify. 2601:646:8E01:7E0B:F88D:DE34:7772:8E5B (talk) 10:48, 10 September 2016 (UTC)[reply]

Lathe. Sounds like everything is on axis. Sounds simple enough (and CNC lathes are less common than CNC mills) that it would be easier manually than setting up the CNC. Andy Dingley (talk) 11:40, 10 September 2016 (UTC)[reply]

Thanks! And will I need to use any hand tools after I'm done with the lathe? 2601:646:8E01:7E0B:F88D:DE34:7772:8E5B (talk) 21:51, 10 September 2016 (UTC)[reply]

You'd maybe use a hand deburring tool, but while it's still in the lathe.

You don't mention sizes, so it's not clear how the outside thread would be cut - either with a die in the tailstock (probably unworkable for a large pipe thread), with a split die box (quickest on a lathe) or by screwcutting on the lathe (most precise). The female thread (assuming it's not too big) would be done with a tap.

You also don't mention how these pipe joints would be sealed. A parallel thread doesn't seal well; usually they'd either be cut with a tapered pipe thread (which can seal when wrapped in Teflon), or they'd be cut with parallel threads purely as a mechanical connection and there would be a turned surface (coned or radial with a soft washer) to make the seal. Andy Dingley (talk) 18:42, 11 September 2016 (UTC)[reply]

The question said NPT, which are tapered. As to the overall task, this sounds like some kind of simple size-changing adapter. If so, the truncated-cone tapering and the thinning to Schedule 40 are probably not needed unless there are some unusual clearance/installation or flow characteristics involved. That is, why can't the inner part of the large male-threaded tube be even smaller (thicker wall than Schedule 40), to the extent that the the female-threaded channel is essentially inside of it? A single tube that has both outer and inner threads, with just enough linear offset so that there is space for a hex area for the wrench. I don't know if the goal is non-standard sizing, or a tradeoff among cost, skill-development, and time, but "stainless steel" "reducing bushing" npt are pretty cheap. DMacks (talk) 21:28, 11 September 2016 (UTC)[reply]

"Why can't the inner part of the large male-threaded tube be even smaller?" Two words: Bernoulli Principle. As for the question of a single tube with both outer and inner threads -- that won't do because the lower part (with the female channel) has to support the catalyst which the upper part (with Schedule 40 walls) contains. In other words, this is NOT a "simple size-changing adapter", but a structural part of a packed-bed chemical reactor -- and the only part which is not available off the shelf (at least not in the size and pressure rating that I need). 2601:646:8E01:7E0B:F88D:DE34:7772:8E5B (talk) 00:44, 12 September 2016 (UTC)[reply]

The threads may require a tap and die set. There is the possibility of machine tapping with a lathe, but this is tricky, as noted at tap and die#Machine_tapping. So, if you don't need to set up a process for many runs, it might be better to do this stage with hand tools (including a vise or two, which would work better while the bar still has a hex shape ). Also, you might want to mill down the hex bar to something closer to a cylinder before working that end on the lathe, as starting from that point could cause unnecessary wear and tear on the lathe, and particularly on the cutting edges. StuRat (talk) 23:44, 10 September 2016 (UTC)[reply]

Tapping in a lathe is far easier than hand tapping. Everything is held square. This assumes you are tapping by rotating the chuck by hand, which may not be possible on a large lathe. You mount the tap in the tailstock, release the tailstock to way clamp and gently bring the tap into contact with the piece. Worth practicing but I don't remember ever stuffing one up. Alternatively on a screw cutting lathe you could mount the tap in the toolpost, and feed it in under power. That sounds like a pain to me, but is essentially how a real tapping machine works. Greglocock (talk) 00:30, 11 September 2016 (UTC)[reply]

Machining, part 2

OK, suppose I have successfully completed the machining operations listed above, but now I have a new problem: the part I've just made is completely covered in cutting fluid (ewww!) If this part is intended for a chemical application where high purity is required, what would be some good ways to de-oil it so that I don't get the cutting fluid into the chemical product? 2601:646:8E01:7E0B:F88D:DE34:7772:8E5B (talk) 10:52, 10 September 2016 (UTC)[reply]

Degreasing tank. Like a dishwasher, with an organic solvent. The best ones are hot sprays, but these have COSH issues, so aren't always hot.

Really clean stuff would be degreased several times in different machines (i.e. cleaner solvents) and might even need to be baked to get rid of chemistry adsorbed into the surface of the steel. Chemistry equipment is easy - high vacuum is much fussier. You might even find yourself using stainless, just because the surface can be cleaner. Andy Dingley (talk) 11:45, 10 September 2016 (UTC)[reply]

Stainless steel is precisely what I'm using, for corrosion resistance. So what kind of degreasing solvent(s) works best for removing the kinds of cutting fluids which are used for machining stainless steel? 2601:646:8E01:7E0B:F88D:DE34:7772:8E5B (talk) 21:53, 10 September 2016 (UTC)[reply]

You can't use the ones that work best! They're a continual struggle between performance and hazard to workers. So most modern ones aren't using the simple, effective, liver-damaging chlorinated solvents that were in use mid-century. See http://www.hse.gov.uk/pubns/eis47.pdf (and similar).

It's now usual for small engineering works to just have small degreasing tanks, used cold, by dip or hosepipe rather than dense sprays or hot vapours. They may use non-chlorinated solvents or even "natural" citrus degreasers. It's enough to clean up scrapyard-condition engine parts for reconditioning, or to clean off cutting fluids well enough for a part to then be coated with lubricant oils or preservatives. This avoids the complexity and air extraction plant needed around the more powerful plants. Those are just found more on larger plants, or those specialising in this sort of high-cleanliness work. For an extra-clean one-off, it can either be done by hand (time consuming for regular jobs) or it can be sent out to a specialist cleaner like [1], who are probably in the general surface finishing trade - hot vapour degreasing is typically a step in a plating process. Andy Dingley (talk) 18:14, 11 September 2016 (UTC)[reply]

So, let me put this another way: of the ones currently available, which ones work best? 2601:646:8E01:7E0B:A013:F78A:60:74D3 (talk) 04:19, 13 September 2016 (UTC)[reply]

Giraffes

Apparently, recent research has determined that there are four distinct species of giraffe, which haven't interbred for over a million years. What I can't find anywhere is whether the species can interbreed. Presumably there have been captive breeding programs that have unknowingly matched disparate individuals? Rojomoke (talk) 11:35, 10 September 2016 (UTC)[reply]

As the new species were previously recognized as subspecies, and identifiable by subtle variations in their patterning, zoos already had a sense whether they were breeding giraffes from different groups. Captive breeding programs have resulted in many hybrid animals. [2] I can't say whether every species has been successfully bred to every other species, but it does seem like some of the new species can likely interbreed. In general, see species problem. Though the ability or inability to breed is often taught as a defining characteristic of a "species", modern research has tended to focus more on the distinctness of two genetic populations rather than simply asking whether they can breed, especially given the potential for some surprising hybrid animals. Dragons flight (talk) 12:04, 10 September 2016 (UTC)[reply]

Well put. A language is a dialect with an army and navy, and it's not too-far off to think that a species is merely a subspecies with the right kind of scientific/political support. It is also generally easier to get public and governmental support for a species, compared to a subspecies. E.g. pygmy chimps didn't get as much ink until they were upgraded to Bonobos with the full status of species. There's a general impression that modern systematics and cladistics are less arbitrary than old-fashioned morphological taxonomy, but there is still a decent amount of subjectivity involved. SemanticMantis (talk) 18:52, 10 September 2016 (UTC)[reply]

I have a friend who is an expert on breeding and taxonomy. Whenever I ask him a question about clades, species, sub-species, etc. his stock answer is "It depends which book you read". That might not be terribly helpful, but I get the feeling it is a truism. DrChrissy ^(talk) 19:01, 10 September 2016 (UTC)[reply]

For one good example of a similar situation, see carrion crow and hooded crow. Genetically, very similar, they are capable of breeding, and there are places where they live together (carrion crows live in western Europe, hooded crows in eastern Europe, and they mingle roughly where the iron curtain once stood). However, they have been reclassified as two separate species, since their offspring lack hybrid vigour and it seems like the birds don't like interbreeding even if it does occasionally happen. You could classify them as one species (and they were, until 2002) but since they isolate themselves, it makes more sense to see them as two separate ones. Smurrayinchester 08:23, 12 September 2016 (UTC)[reply]

The original study says that Concordance between maternally inherited mitochondrial and biparentally inherited nuclear markers indicates reproductive isolation for at least four giraffe groups. This lack of gene flow is unexpected, because wild giraffes are highly mobile [19] and can interbreed in captivity [20]. Basically as Dragons flight said. Henry ^Flower 20:06, 10 September 2016 (UTC)[reply]

We humans like to put things into convenient boxes (classifications), but the real world isn't that simple. Ring species is interesting. Dawkins writes about it in Ancestor's Tale. 86.20.193.222 (talk) 04:09, 11 September 2016 (UTC)[reply]