Metal fatigue and clock making

[Clive CasselParticipant @clivecassel81348]

Mild steel clock frame – how to overcome metal fatigue

[Clive CasselParticipant @clivecassel81348]

I’m making a John Wilding 16th century style clock which requires a frame from 1/8″ x 1″ ms bar.  The frame requires right angle bends in the plane of the steel bar. I have tried steel bar from two suppliers and both bars snapped at some point. The method in the handbook says:

 

“A saw cut is made to a depth of about three quarters of the thickness of the strip. The cut is now closed up by bending the strip … A second saw cut is now made in the same position removing the metal which has been pinched together …the cut is closed up by bending … and the process is repeated until a right angle has been formed.”

 

On each occasion the bar has snapped at the joint before completion. What to do?

 

Many thanks in anticipation

 

Clive Cassel 

[David Clark 13Participant @davidclark13]

Hi There

Heat it to red hot and then bend it.

regards David

 

[Jeff DaymanParticipant @jeffdayman43397]

You might consider mitre cutting four pieces, welding them, grind the welds out and radius the outside corner. If an inside rad is desired for looks, an infill piece with the right radius could be welded in also. Electric arc or MIG welding would be best for this, as the filler wire is the same material as the stock, for best (most invisible) appearance for later operations.

 

Another option would be to have a picture frame shaped part cut by laser, waterjet or flame, then clean up the machine-cut egdes with files and grinder.

 

For an antique look, you could heat up the parts in a fire a few times to red heat, which will blister the steel a little and round off the sharp edges, as well as soften and stress relieve the steel. In the 16th centrury the iron used for such a clock frame would certainly have been fire-worked and very ductile.

 

JD

[Nicholas FarrParticipant @nicholasfarr14254]

Hi, another technique you could try is, once you have made your saw cut, you could use a three corner file and widen the saw cut out to a V shape of say about 95/100 degrees inclusive. This would have the same mitre effect that Jeff has mentioned.

 

Wrought Iron was probaly used in the 16th century, and like Jeff says it was more ductile and less prone to work hardening which will lead into cracking and snapping, which is what your MS is doing. Wrought Iron. is not very common off the shelf these days.

 

Regards Nick.

Edited By Nicholas Farr on 19/12/2010 15:48:24

[Gordon WParticipant @gordonw]

I don’t understand this, are you trying to bend across the 1″ dimn. ? ie. cut 3/4″ deep. If so mitre end and weld as said. If bending the 1/8″ thickness can do it cold with a big vise and square block and hammer, can get a good sharp corner if needed. Wrought iron is not made anywhere in the world anymore, the last lot I got, 20 yrs. ago was some re-worked stuff from an industrial museum, but I would like to be proved wrong.

[NJHParticipant @njh]

Hi Clive

 

I think David has it right here. MS would bend fairly easily if you can get it to red heat – however I don’t know what facilities you have for heat treatment. Whilst the welding method would, no doubt, achieve the desired shape it seems to me rather out of sympathy with 16th century clock making !

You could drop a mail to John Wilding ( via http://www.ritetimepublishing.com) and ask him for the source and spec.of his material.

Quite an interesting clock  ( but not a reliable time – keeper I suspect!) . How about a few photos as you progress the project?

 

Regards

 

Norman 

 

Edited By NJH on 19/12/2010 16:17:14

Edited By NJH on 19/12/2010 16:19:40

[Keith LongParticipant @keithlong89920]

Don’t know about welding being out of sympathy, I’d have thought that forge welding it from strip was quite probably how they’d have done it back then. OK mma or mig wasn’t around but it would be in a similar spirit. 

Quite possibly they’d use a combination of the two techniques, cut and bend and then hot forge the joint solid. 

Keith

[TerrydParticipant @terryd72465]

Hi Clive,

 

Are you using bright drawn MS or black hot rolled bar? and when you say ‘in the plane of the steel bar’  I presume that you mean you are bending the 1″ wide section so that the frame lies flat in that section, if you know what I mean. What radius is the bend? Is there a picture or diagram you could post?

 

Terry

 

PS I just reread your post after being sidetracked by another posting.  I have found an image of the clock and see what you mean.  Hot rolled bar is more malleable than bright drawn and David is correct in that if you can get localised red heat and use bending bars or similar in a vice you should be able to bend without saw cutting.  That would be how a blacksmith would do it so is in the spirit of the original

 

T

Edited By Terryd on 19/12/2010 19:06:25

[macmarchParticipant @macmarch]

If you are not going to weld it, then I would suggest using the “black mild” stuff that B & Q sell.  It brazes ok, machines to an acceptable finish and bends to quite a tight radius. It does need a little heat if bending tight.  The surface coating splits off as you bend it.

 

I have checked with many sources for a ductile bending quality steel but it seems that there is no modern spec. unless someone knows better! 

 

chhers

Ray

[NJHParticipant @njh]

Hi Keith

 

Yes on reflection you are quite right – forge welding was correct for those times. I guess I was distracted by reference to Arc & MIG welding  – both black arts to me!

I the case of this clock however just a bit of heat to bend the steel is all that is required. As stated above though the black variety is the one to use.

 

Cheers

 

Norman 

[TerrydParticipant @terryd72465]

Hi Norman,

 

The problem with bright drawn steel bar is that the outer skin is stressed and work hardened by the manufacturing process.  By cutting 3/4 of the way through you are trying to bend a less malleable and ductile skin. Hence it cracks and fails. which is why we suggested black bar, it doesn’t have the built in stresses and work hardening.  But there are other interesting ways to make your frame.

To make a really sharp right angle instead of bending you could try a dovetail. 
In metal it can be made in a similar way to a wooden dovetail used in
furniture making and then silver soldered, and even soft solder will
give a very strong joint.  Flux before assembly though. Also it doesn’t need to be quite as accurate as a wooden variety as you’ll see below.

 

It can also be made to lock tightly together to make an equally strong joint without any further soldering etc and this would be quite an authentic method to use.

 

In this latter method you cut the single tail and pins  as before but a couple of mm too long.  
The tail is made first and the pins marked from it as in woodworking
practice.  The most important part is to file the shoulders accurately
and square, take great care over this.   Now a secondary angle is filed in the vertical cut out of the pin,
undercutting  them.  The joint is assembled and the end of the tail is
peened so that the material is forced to fill the ‘undercuts’ in the
pins, locking and securing the assembly.  The pins can also be peened in
thus taking up any inaccuracy and the excess material filed and tidied
up.  It almost takes longer to describe than do.  The only tools you
need are a junior hacksaw, 3 square file, a ball peen hammer and a vice
(and basic marking out tools of course).  I’ve included a few diagrams below to try to illustrate better what I mean.  It’s worth trying on some of your scrap and it can be done with the material you already have.

 

 

1: Typical plain dovetail ready for soldering:

 

2 Alternative method with no soldering etc needed – Pins formed by sawing and filing, then a secondary angle is filed as indicated by the shaded areas to undercut the pins:

  3: Joint assembled showing gaps caused by undercutting pins:

 

 4: Final Joint.  Tail has been peened to force material into gaps (shaded areas) thus making the joint rigid and secure without soldering or welding.

Terry

Edited By Terryd on 19/12/2010 23:00:28

[Keith LongParticipant @keithlong89920]

Hi Terry

I like that idea – seems right and “country-style” somehow. If you had a bit of relief filed on the upper edges of the “tail” you could swage the pins over as well make it virtually impossible to pull apart.

If they could build the “Ironbridge” with “woodworking” joints a clock should certainly work.

Keith 

[TerrydParticipant @terryd72465]

Hi Keith,

 

Even with the method I’ve outlined the joint cannot be
pulled apart as you have a double wedge effect.  It really makes a
strong joint. Of course the joint can be milled using a dovetail cutter
but I like to make it with saw and file, it’s a bit more challenging. 
It’s a very old technique and the old blacksmiths would cut the joints
quite roughly with a chisel on hot material and forge. it together wile
red hot.  Also to improve the look and make it more decorative, the pins
and tail would not be filed down, instead they would be planished when
cold with the round peen of the hammer to give a rounded ‘hammered’ look similar to a coppersmith on fire hoods and table tops etc.

 

Properly done it is a very decorative and authentic method of this type of joint.  I use the soldered variety for making upstands for some of my small oscillating steam engines.  It makes a change from rivets and belongs to a time before screw threads were invented.  I only file the shoulders of the joint.  I try to leave the other cuts as sawn as when planishing, the material is forced into the marks left by the saw, strengthening the joint.

 

By the way, it was also used by the old
woodworking plane makers in the time between the wooden plane and the
modern cast body plane.   The plate sides would be attached to the sole using rows of dovetails and then this body infilled with decorative hardwoods with brass and steel fittings.  They were works of art.  If you look very carefully you can just about make out the dovetails joining the side plate to the sole on this example.

 

 

Terry

Edited By Terryd on 19/12/2010 23:39:18

[TerrydParticipant @terryd72465]

Hi There,

 

Sorry to hog the thread but my first post on the dovetails should have been addressed to Clive rather than Norman, blame the late hour (no alcohol to use as an excuse either).  I apologise for the mistake.

 

Terry

[John OlsenParticipant @johnolsen79199]

Now that the original problem has been pretty well dealt with, can I put on pedantic mode for a moment and mention that the original fractures had nothing to do with fatigue. They would be a normal fracture caused by stressing the material past its elastic limit, and would have the typical chrystalline appearance of such a fracture.

 

I don’t know if modern theories have changed, but our mechanics and strength of materials lecturer made a distinction between fatigue, as commonly seen in materials like aluminium, and crack propagation, as often seen in highly stressed components like steel connecting rods. The way he explained it, fatigue occurs with loads that reverse, and the time it takes for it to become significant depends on the magnitude of the load and the number of load reversals. So far as was known at that time, there was no lower limit although with a small enough force the number of reversals required would require such a long time that it could never actually be acheived. Not all materials are prone to this type of fatigue, aluminium of course being the classic example. One of his examples slides was a photograph of a wing spar from a Bristol freighter that failed in the air over Christchurch (NZ) The effect of fatigue of this sort is that otherwise strong enough material loses its strength, and cracks will begin to grow, leading to failure. The only way to avoid this process is to keep a record of the usage, and discard the component when it has reached a known maximum safe life.

 

On the other hand, according to him, crack progagation can occur in pretty well any material that is not too ductile. If the material is very ductile it will just give a little at the stress concentration, releiving the local stress without extending the crack.  With less ductile material a microcrack can grow at a local stress concentration and keep growing each time the stress peaks, eventually getting long enough that failure occurs. Because the mechanism is different, it is possible to stave off the begining of this process by means like polishing the surface, to remove stress concentrators like scratches. Also you can test a used component, and if no cracks have started you can reuse it, as used to often be done with connecting rods.

 

Of course, with either mechanism once the cracks have got going, the end result is inevitable and predictable, and the resulting appearance of the failure is likely to be similar. So what I am curious to know, if there are real metallurgists  here, is this distinction still made? Or was it just a bee in the bonnet of that particular lecturer?

 

regards

John

[TerrydParticipant @terryd72465]

Hi John,

 

No one has mentioned ‘fatigue’, but we have mentioned work hardening.  This occurs when any metal is stressed by physical means, whether that is fatigue caused by load reversal in aircraft spars or in con rods, or distortion by hammering into shape etc.  

 

I’m no metallurgist but I seem to remember from my Mat. Sci. lecturers that any work hardening distorts the crystals in the metal.  As they are distorted the atomic bonds are stretched and become less able to become further distorted so the grain boundary between the crystals become more distorted and stressed as they have to sustain more distortion with further working.  And it is the grain boundary , not the crystal which fails (the atomic bonds in the crystal are too strong to fail under normal loads) as it has become distorted and over stressed.  This then increases the strain (and stress) on the remaining boundaries, the process is accelerated and the material eventually fails, final failure occurring suddenly as many boundaries fail quickly.  This was epitomised (exemplified?) by the failure of an overhead walkway in a shopping mall in the USA when the supporting rods failed due to over stressing caused when the builder deviated from the design ( a case of a craftsman thinking he knew better than the designer!).  Quite a number of people were killed.

 

If we normalise the component with heat , the heat relaxes the crystal boundary and the slow rate of cooling allows the crystals to relax into their original form. and the  stresses in the grain boundary reduced (or removed).  Of course if we overheat the metal the grain boundaries become so relaxed that they allow the crystals to flow and the metal enters it’s liquid phase.

 

As Bright bar is drawn down through the dies, the surface of the bar becomes work hardened due to the crystals becoming deformed during the process.  There is less distortion in the centre and there is an inbalance in stress concentration at the grain boundaries through the material,  hence when machining Bright drawn bar it is best to remove equal amounts from opposite surfaces in order to balance the stresses otherwise distortion of the machined component can occur.

 

In Clive’s clock he is cutting through the bar leaving only about 0.8 mm of material to bend and if he is using bright bar which is not normalised he is trying to bend the most highly work hardened and stressed part of the bar. hence the grain boundaries will fail at the outer surface thus propagating a crack and finally inevitable failure.

 

I believe that your lecturer was correct in saying that cracks can be propagated at stress concentration points like scratches and other faults when a metal component is loaded, even with a constant unchanging load and that these can be minimised by polishing to remove or reduce such faults, but that is not the only situation where cracks can begin.

 

Now I am sure that someone with more knowledge than myself will be able to shoot me down and add to the debate.

 

By the way John, it is a treat to see someone who can structure a piece of writing using proper paragraphs. 

 

Regards

 

Terry

[RJWParticipant @rjw]

Terryd, re: Fatigue, it Was mentioned actually, please refer to the actual topic heading!

Great thread all the same.

Regards,

John

[Gordon WParticipant @gordonw]

Still can’t find the original drawings, but- corners like these, in the olden days, were often made by upsetting the ends and forming into 90deg. “brackets”, then riveting onto flat pieces. An easier modern way is to make the saw cut on the outside of the bend, then weld, makes a neat joint. The old fire welding blacksmiths would have used electric welding if they had the chance. Sorry not properly formed para. but have found the spell -check button.

[TerrydParticipant @terryd72465]

Hi RJW,

 

Of course you are correct and I apologise to John Olsen if I have contradicted anything he said, I was just trying to extend his excellent contribution.

 

 

Hi Gordon,

 

Only one paragraph needed, . 

 

Best regards and seasons greetings,   (we need a holly icon in the smileys!!)

 

Terry

Edited By Terryd on 20/12/2010 10:41:49

[Ian S CParticipant @iansc]

   As an aircraft engineering apprentice i was very soon told the reason for not using “Aviation Snips”, invented before the war in America, and after a while the began to wonder why aeroplanes/ airplanes were falling out of the sky, if you look at aviation snips you will see that the edge is finely serrated, thats the answer.

 

   One of our  instructors at NAC (now Air New Zealand) was on the next green , on the golf course when the 10,000 pop rivits ceased to be flying in close formation, as the Freighter crashed he was one of the first on site.  The aircraft was designed toward the end of the war, and was virtually a powered replacement for the transport gliders, and giving components a life in hours flown was not thought necessary, disposable aeroplane. They actually gave good service.

 

   On our Cessna aircraft we had a number of fatigue problems, the Cessna 185, a tail dragger used to break the tail spring (to which the tail wheel is attached), some one came up with a stronger one, but none of the pilots would fly a plane with it fitted, they said rather the spring breaks, that the tail falls off.  Ian S C

[Nicholas FarrParticipant @nicholasfarr14254]

Hi Terry and John, the point about about stress consentration points is a valid one. During my colleage welding course work we had to cut a 1″ section at right angles to a butt weld using a low heat and non pressure input cutting technique, i. e. a hacksaw. this section was then given the bend test, which was a U bend round a round former. I can’t remember the ratio of the former diameter to the plate thickness, but it was about 25mm for a 3mm thickness. The weld bead had to at the very bottom of the U shape. The test piece had to bend without breaking or any sine of cracking to pass as being a sound weld. The test piece preperation was to file tyhe weld flat to the plate and along the egdes over a longer length to that which the bend required, this also included rounding off the edges and ensuring that there were not even the smallest nicks present. Scratches, however in this situation was not a real concern. It was proved that sharp edges and hacksaw marks propagated cracks and falure by taking two test pieces from the same section of the same weld and preparing one as specified and one leaving it with sharpe corners and small hacksaw marks and while the one correctly prepared passed the test the other produced cracks or even fractured, this was of course if your weld was a good one in the first place.

 

Fractures do indeed occur along the grain boundry. The term normalising is usually refered to the heat treatment that gives a metal the finest grain structure possible and is subject to a control temp/time and metal content ratio. If a metal is heated above or longer than this critical area, grain growth forms. The larger the grain is, the weaker the metal is.

 

Bright steel which has been cold rolled has an elongated grain structure along the path of the rolling process. The lenth of elongation deminishes to is central section, which as Terry says is stressed and as you will find that if you machine off one side it will bow away from that side. The amount of bowing will increase the more you remove from the same side. Even if you clamp the piece down to prevent it bowing, once it is released it will bow. The only way to prevent this happening is to stress releave it prior to machining. But normalising will give it ultermet strengh.

 

Normalising shouldn’t be confused with annealing.

 

Regards Nick.

[RJWParticipant @rjw]

Hi Terry, unfortunately my posting got cut short due to SWMBO
demanding my immediate presence, so my apolologies if it came across ‘a
bit short’, couldn’t get the ‘smileys’ to work either :>(

 

The
‘Fatigue and clockmaking’ header drew me to the topic
on two counts, one; because I’ve been beavering away for the thick end
of 2 decades restoring Antique clocks and pocket watches, and secondly,
because the word ‘Fatigue’ in relation to metal, always brings to mind
an ammusing interlude during a very intense study session in the 1990’s,
which has never left me to this day – especially around aircraft!.

 

I’ve
also witnessed many flame wars on various ‘Horology’ forums (fora …
forii?) especially, concerning Ammonia solutions used in cleaning brass
causing fatigue fractures, and wondered initially if this topic was
going that way, thankfully not (I hope)!

 

 As
regards John Wilding’s clock design, it depends upon whether the builder
(or Clive) wants to faithfully and rigidly replicate the Wilding design, or use it
as a sound and proven basis upon which to build in modifications to suit the builder, thus stamping
his or her own signature on the piece.

 

My own inclination would be to use  joints similar to those which Terry illustrated on the frame rather then to bend up bits of metal, although I would probably tend to make them even simpler, by using squares rather than dovetails (the KISS principle).

 

It
might also be worth remembering that Blacksmiths were making clocks
long before ‘Horologists’ got in on the act, and in fact, the once hallowed 
‘Clockmaker’s Company’ grew in the first place,  from the ‘Blacksmiths
Company’.

If some other method of jointing is desired, it might be worth additional research of the methods used back then by studying some typical iron movements.

 

I have a couple of French Comtoise or Morbier
clocks, both of which use iron frames, one uses pins on the uprights
rivetted into the top and bottom plates, the other simple squares
rivetted into sockets.

Unfortunately, both clocks are in France and out of reach until April, so I can’t add photo’s.

 

Although some well founded theories have been expounded here on metal Fatigue, and arguments against the term ‘Fatigue’  in Clive’s particular intance, the fact remains that however and whatever the cause of a piece of metal breaking, most people will Still refer to such a break as being caused by ‘Metal Fatigue’ ……. even I do, lets face it, it’s a damned sight easier expression!

 

Regards,

 

John

 

 

[RJWParticipant @rjw]

Ian C, I think there must be a tad of aeronautical telepathy happening here <LOL>

 

A bit of light relief here I hope, and ever so ‘not quite’ off topic ………… ;&gt)

And for the pilots of you out there, you may never look at an
aircraft wing again in flight without a smile creeping across your face.

 

In the 1990’s, having seen the writing on the
wall for the end of my engineering career, (busted back and worn out
joints)! I decided to ‘upgrade’ my private pilots licence whilst I could
still walk, and become a professional flying instructor – which meant
by then (thanks to the CAA) gaining a CPL first.

 

The ground school element for the written papers
was for me,  a full time residential course for each discipline – Aircraft
technical and Navigation subjects, not much in the way of light relief,
and little short of a gruelling  war of attrition for many students – financially And accademically!

There were though, a few memorable occasions which will stick in my mind forever,  and one such event was during a discussion regarding metal fatigue in airframes.

 

For the Aircraft Techinical lectures, we had a
tutor who was a brilliant mathematician, and whom I vaguelly remember
had something to do with Helecopter design, maybe rotors, but he knew
his stuff inside out and backwards, and as such wasn’t often questioned, and whatever he said, we tended to hang onto his every word.

NB. If you’ve ever forked out serious money to write CAA exam papers that you seriously didn’t want to fail,  you’ll know the feeling!

 

Anyway, one day aircraft cabin pressurisation and aerofoil design (wings)
and life cycle fatigue were being discussed, (and I’ll wager that not a pilot in
existance hasn’t at some point, looked at a wing in flight and secretly
prayed the bugger wasn’t going to fall off) so he had our rapt
attention.

 

The chap turned to the white board and calm as
you like, pointed to a diagram of a typical airliner wing exclaimed
…….. ‘I don’t know why designers don’t just drill a line of holes straight
across the stress points and have done with it, the wing would probably never
break there and it’d save them billions!.

 

Well, we
all sort of sat a bit nonplussed as we digested this gem of wisdom, and
waited for the Guru to expound  further, but the pregnant pause got too much for one student, who took the bait and had to ask ‘Why’ ………… the quick reposte he got and with a dead pan expression was ………..

 

‘When was the last time you tore a piece of paper off a bog roll, that actually tore on the perforations!

 

Ok, we were all stressed out of our brains and bank balances with impending  exams, and maybe it was funnier then than in the telling now, but we all just fell about howling with laughter, and coming from such a noted ‘boffin’ who was as dry as toast, it was a magical moment I can tell you.

 

Well, I got my CPL, then went on to become a
commercial flying instructor, and latterly a flight examiner, and I can
tell you with my hand on my heart, there was never a moment from then on, especially when entering or recovering from a steep turn, power on stall, or a fully developed spin, that I didn’t look at my aircraft wings and remember the bog roll reference without a smile on my face, or occasional chuckle!

Invariably during those recoveries, my students would gaspingly ask what the hell I was smiling about, and recounting the tale, would see them laughing nervously ……….. as they surreptitiously scrutinised the wing on their side whilst making out they were doing a ‘look out’ ……………… :&gt))

 Maybe they were looking for a line of holes <LOL>

 

Regards,

 

John.

 

 

 

[TerrydParticipant @terryd72465]

Hi John (RJW),

 

Actually, dovetails are no more difficult to cut than the plain joint you described.  it just takes a little bit longer to mark out. I urge you to try it on a couple of scrap pieces  ‘cos if you did you will wonder why you ever thought it might be difficult.  It doesn’t even have to be super accurate, just so so,

 

Best regards and seasons greetings

 

Terry

[Author]

Posts