hydraulic valve - opens on set pressure, closes no pressure

I don't entirely follow that paragraph. See my reply to Richard.

Re a device with memory, a directional valve with two detented solenoid actuators would do the trick. One solenoid controlled by a pressure switch that closes at the high pressure setting, the other by a pressure switch that closes at the low pressure setting. Or the hydraulic equivalent.

formatting link

Reply to
Ned Simmons
Loading thread data ...

Stunning detail. I see that with

  • test method and hardware
  • control and data acquisition
2million cycles would be a start. With well-performing welds, you'd want to get out to 20million cycles or more (200Million is something around what a ship or bridge gets in its entire period of service, but can be a bit of an ask in test). I've programmed in Basic. Access to devices / ports - hadn't thought about it being that convenient and tailor-made.
Reply to
Richard Smith

Ned, Jim, everyone - massive thanks. Rich S

Reply to
Richard Smith

----------------------

I can't claim credit for the idea, it was the engineer's suggestion when we (Unitrode / Texas Instruments) wanted to design and operate evaluation boards for new user-configurable ICs from customers' unmodified laboratory computers, which at the time were typically former office desktops and laptops running Win98. We started with Visual Basic but quickly found that it lacks the hardware control instructions of QBasic, and Windows polls the printer port to detect newly attached devices. DOS + QB give full read/write access to the I/O address space, interrupted only to update the clock.

The only change to the computer was setting the BIOS to boot from a DOS floppy if present, else to Windows. The same can be done with a USB flash drive using HPUSBFW. FAT32 USB flash drives are big enough to store programs and large data files without access to the NTFS internal drive though older FAT32 hard drives could handle either DOS or Windows, up through XP.

formatting link
The computer boots normally with the flash drive removed and you can read any data log files the QB program created, such as cycle intervals which might increase when the sample began to stretch, and indicate the point of failure if you can't otherwise sense it.

I had previously assembled one-time computer to hardware interfaces with added plug-in boards, a purchased digital I/O card and a 16 bit A/D converter for the Macintosh that I designed. The printer port and DOS/QB method turned out to be easier for relatively simple tasks. I'm also very familiar with relay ladder logic controls if you choose to go that way.

Do you have the equipment or machinist friends to consider custom machining as part of solutions? I couldn't do nearly as much without my lathe and milling machine.

Reply to
Jim Wilkins

Machining, making parts and equipment... My cider-maker / cider supplier could feature in that... Can see a need to trade favours...

Digression into computing. I often put "emacs" on a CD-ROM, take it by IT-support and invite them to check it (fully up-to-date virus check software, etc, etc.) If it's a write-once CD-ROM, it can't be altered ever again. So if they trust that is all I will ever put in the drive of "my" networked computer, that's sound in terms of IT security.

The reason for this is; I have a range of suites of functions for various engineering tasks which run in the test-processor I am using now - the famous / well-known "emacs". They will inject the answer straight into the document you are writing. You can quote the function and the values you fed to it, and the answer it gave. Complete record. "sketching" your way to often high-value decisions. Plus I do most other text-based thing in emacs. For examples I didn't just type " Thermo-Mechanically Controlled-Processed High-Strength Low-Alloy " I typed tmcpqc hslaqc When you are typing documents with huge strings of standard quotes of Standards, Company Specifications, etc, that can save a huge amount of time and effort, and leave your mind clear to think of the big picture. Boot off a CD-ROM? Could do?

Reply to
Richard Smith

Hi Ned.

Your real-world experience enables you to answer this question?

You've already raised the point

"... Whether this is a practical way to control your device in the real world is another matter."

I was thinking of constancy of pressure at which it triggers. This thing is perfect when sequencing. For a fatigue test, hydraulic pressure at trigger = peak force. This is crucial to the test - for the peak force to be known and constant. Would this device settle down to triggering at a very constant pressure, do you reckon? Within reason, so long as it stays constant, what that pressure is is just fine - plot the "F=PA" force on the "S-N curve / plot" for comparability of data.

(That's why I designed a parallel small system "reference pressure" concept - so the peak force / pressure is accurately at, but neither more or less, than a constant reference pressure)

Rich Smith

Reply to
Richard Smith

"Jim Wilkins" snipped-for-privacy@gmail.com writes: ...

Machining, making parts and equipment... My cider-maker / cider supplier could feature in that... Can see a need to trade favours...

----------- So not easily, perhaps shortening a bolt but not to the full custom extent I would otherwise suggest, like the small piston in the cylinder end cap. Experimenting is difficult when you are limited to only what you can afford to buy. My shop is what an inventor would have dreamed of in 1960 though perhaps not today, after 50~60 years of wear.

How about electronic test equipment? Capturing the brief peak value from an analog pressure sensor during rapid cycling will be difficult without a digital storage oscilloscope.

formatting link

Reply to
Jim Wilkins

... This is crucial to the test - for the peak force to be known and constant. Would this device settle down to triggering at a very constant pressure, do you reckon?

-----------------

I don't know about that valve but I do have lots of R&D experience on difficult projects.

Generally you design and build what you predict (hope) will work, then measure and correct until you run out of time, money or patience, and declare it "good enough". The limit is how accurately you can measure.

If the valve is inconsistent you may simply need a better filter, or different oil viscosity. A recording of cycle intervals and peak pressures would show you if a problem develops and perhaps hint at why, i.e. was the change sudden or gradual.

I save measurement data into text files that can be loaded into a spreadsheet as *.csv for analysis.

Reply to
Jim Wilkins

_____________ Can do easily.

Hit the key during startup that enters the BIOS setup and put the CD-ROM ahead of the hard drive in the Boot Sequence. On my Dells it's F2 before the self-test completes. F12 lets me select any bootable source for this session only.

My Lenovo laptop is a little different. It came in a Fastboot mode that bypassed the chance to enter the BIOS.

formatting link

Reply to
Jim Wilkins

I've designed hydraulic systems that included sequence valves, but I wouldn't want to make a prediction on how repeatable the cracking pressure might be, especially at high cycle rates relative to the size of the device. It's also likely to drift with oil temp, though that's a slow-changing parameter that could be compensated for manually. Probably best to speak to manufacturers, Eaton/Vickers, for example.

Reply to
Ned Simmons

It's easy to run out of all three quickly fiddling with a new hydraulic design. They're messy, difficult to change, and often don't allow easy access to the parameters you'd like to measure when debugging.

Reply to
Ned Simmons

Ned - that's what I "feared" - it will likely settle down to a stable pattern, but what that is will drift around in time, with many variables, including ambient temperature, etc.

Running 24hrs a day for many days, you couldn't keep fine-tuning to stay on one condition.

I'm seeing why servo-hydraulic with digital logic control on the basis of pressure transducer(s) is the way things are done.

By the way Ned - "drift" affects welding - not sure how much that is your thing. "drift" is central to a funny (?) story.

This Company owner knew I wanted to get ahead, so ordered the Foreman

- "Don't let Richard weld - he doesn't know how to". Well, this job comes in with far thicker plate than anyone has ever met - even the "old-timers" with 40+ years experience. They couldn't get their welds to "stick" in those slot-welds. None knew spray-transfer. Co. had one really good machine, and I flipped it up into spray and was putting in these slot welds easier than shelling peas. These big structurals were up on plinths, so I was up there in the middle of the shop, on top of these things, putting in the welds no-one else could.

No-one came and asked me how I did it.

Reason - they knew I'd say "Don't know. I can't weld. I just pick up the torch and it seems to work".

The reason they couldn't just copy my setting, or order me go buy striped paint while they take the torch and continue on that setting is - conditions drift... You have to tune and get back to the right condition every few minutes. Variables? Line voltage? Ambient temperature? How long since the machine was turned on? etc.

If you drift downwards, the wire will stub and the weld will transition to dip transfer. Get too long and it can burn-back to the torch which will instantly destroy the contactor tip and probably the shroud. You'll spend a long time rebuilding your torch (N.Am. "gun") if you don't know how spray works.

So yes, drift...

Again Ned - thanks, appreciated.

Reply to
Richard Smith

Could be more than that... Device with an eccentric crank? Simple and presents a huge bearing area to take high forces.

Reply to
Richard Smith

There's a mini-computer called a "Raspberry Pi". Designed by group endeavour of hobbyists. Meant for education. Uses a "merchant" embedded-chip for eg. washing machines as its processor. Might look if that offers a way. Lots of devices been matched to it.

However, note you "PC architecture / QBasic" way.

Reply to
Richard Smith

It's easy to run out of all three quickly fiddling with a new hydraulic design. They're messy, difficult to change, and often don't allow easy access to the parameters you'd like to measure when debugging.

Reply to
Jim Wilkins

There's a mini-computer called a "Raspberry Pi". Designed by group endeavour of hobbyists. Meant for education. Uses a "merchant" embedded-chip for eg. washing machines as its processor. Might look if that offers a way. Lots of devices been matched to it.

However, note you "PC architecture / QBasic" way.

---------------------------- Sure. The Pi has the advantage that you can buy and plug in an A/D converter to read the pressure sensor, and a keyboard and display to monitor the test, no technician bench skills required. I haven't played with one and don't know the learning curve for it. Are you generally familiar with computer hardware and software architecture?

As for the eccentric, I considered it because you could vary the piston stroke, but you'd need a lathe to make it, the pump drive is more than a friend could whip out in a spare hour. A pump built into the cylinder end might give the fastest cycle rate because there's no flow restriction. You could fine tune the peak pressure during operation with a screw that displaces oil. A cheap used tie-rod cylinder with a scratched bore could be cut down to be your pressure chamber. Cylinder rebuilders can provide the tubing in any length.

On my tractor's homebrew bucket loader attachment hydraulics I turned the head of a bolt round and grooved it for an O ring, so it can screw into or out of the oil space without leaking. It operates the variable pressure relief valve I made to replace the fixed relief the control valve came with. The tractor's front tires turned out to be the weakest link that limited how high I could set the pressure.

The decision comes down to what you can build or buy. I've spent significant time and money becoming able to build what I or the customer wanted, electrical, optical and mechanical. Right now I'm upgrading my sawmill and its overhead gantry hoist to handle a larger log than it was designed for.

The sawmill is a large horizontal bandsaw made from salvaged motorcycle wheels. Thursday was lost to the miserable task of prying off the drive wheel's tubeless tire, scrubbing the corroded bead seating surfaces and getting it to hold air. The slow air leak reduced the blade's initial 500 lb tension and let it wander in the >20 inch long cut. I learned manual tire repair at the Hardway School (an Army motor pool) on split rim truck tires. As the tire store manager told me, a 2" right angle air sander is the perfect tool for cleaning up a leaking rim. My tractor's rusted wheels will be its next victims.

Would fatigue cracking in oil be similar enough to cracking in air, which oxidizes freshly exposed steel?

Reply to
Jim Wilkins
<snip>

And what about heat? Flexing metal creates heat. If you accelerate flexing to speed up failure detection you will likely create heat that would not be present in its actual use...

Reply to
Leon Fisk
<snip>

And what about heat? Flexing metal creates heat. If you accelerate flexing to speed up failure detection you will likely create heat that would not be present in its actual use...

Reply to
Jim Wilkins

- "Don't let Richard weld - he doesn't know how to". Well, this job comes in with far thicker plate than anyone has ever met - even the "old-timers" with 40+ years experience. They couldn't get their welds to "stick" in those slot-welds. None knew spray-transfer. Co. had one really good machine, and I flipped it up into spray and was putting in these slot welds easier than shelling peas. These big structurals were up on plinths, so I was up there in the middle of the shop, on top of these things, putting in the welds no-one else could. ...

--------------------------

You could get a high-paying job welding submarine hulls, and be invited aboard for the test dive.

Reply to
Jim Wilkins

There is a submarine base nearby. Very exacting recruitment.

Reply to
Richard Smith

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.