Aluminum Work Hardening 5

[AlanD]

I've been involved in a long-term debate with people who in my opinion do not have the technical background, over whether Aluminum mast (for sailing) will work harden without the masts becoming permanently bent. The masts are a tubular section of either 6061 or 6063 grade and heat-treated to T6

From my knowledge of dislocation movement work hardening will only occur when the material yields. But below the yield point, the dislocation is unimpeded.

The masts do flex, but it seems wrong to me to think that they work harden while they flex. To make things more difficult, the mast suppliers are now labelling the masts to indicate that new masts should not be used in strong breezes, until the masts have been "broken in".

Does anyone have any comments? I'm particularly interested in any papers written on the subject.

 

[etch]

Perhaps the more likely term should be aged harden. Aluminium with a Mg content approx above 0.5% will age and become stronger through time, typically components made of this material are laid down for a period and allowed to age before eing put into service. One way is simply to conduct hardness tests every few days untill a stable result is obtained.

 

[AlanD]

I don't think that is likely. The alloys mentioned are solution treated and artificially aged (T6 temper). The artificial aging temperature from my understanding is between 160-175°C (approximately 310°F. If the alloys are correctly aged, the aluminium should be in the fully aged condition.

At room temperature these alloys shouldn't be naturally aging to any significant extent and if they did they should become overaged and hence softer.

 

[TVP]

It sounds as if the people with which you are discussing this issue do not understand Materials Science and Engineering. Work hardening, more properly called strain hardening, occurs when a material is plastically deformed, i.e., when stresses beyond the yield strength cause dislocation motion. In the case of a sail mast, the forces should be causing only ELASTIC deflection/deformation, and not PLASTIC deformation. Strain hardening most certainly does not occur under elastic loading conditions.

The following websites have some additional information on the subject:

http://www.tpub.com/doematerialsci/

especially this chapter:

http://www.tpub.com/doematerialsci/materialscience59.htm

http://www.virginia.edu/bohr/mse209/class.htm

especially this chapter:

http://www.virginia.edu/bohr/mse209/chapter7.htm

For an expert-level discussion, you should refer to R.W. Hertzberg's excellent book Deformation and Fracture Mechanics of Engineering Materials, available from Wiley using the following link:

http://www.wiley.com/cda/product/0,,0471012149,00.html

 

[CoryPad]

I don't think I agree with the people to whom AlanD refers, but I will play devil's advocate here.

Since masts are long tubes, the stresses are concentrated at the base. This also concentrates any deformation. Is it possible that initial loadings cause small deformations concentrated in this area. This may not cause gross mast shape change, but may cause sufficient strain hardening to prevent more deformation from future loadings. One thing that seems to disprove this idea is that 6061-T6 has little strain hardening capacity: yield strength = 275 MPa & ultimate tensile strength = 310 MPa. This small strength change wouldn't allow significant improvement.

 

[kenvlach]

Instead of speculating why sailors think masts need breaking-in, I visited boaters' sites including a forum at

http://www.briontoss.com/wkstone-0....ntoss&Profile=message_board/showMessages.prof

I didn't find the exact Q & A, but I drew the following conclusions:
1) Sailors are definitely not trying to work-harden their masts by plastic deformation. It is much trouble to straighten a bent mast and then fit a concentric sleeve over (or inside) that section for reinforcement.
2) They know a lot about the metallurgy of Al 6063-T6, how welding weakens it, galvanic corrosion, anodizing vs. polyurethane coating, etc. They know some engineering failure modes, and space and size holes to avoid stress concentration. Also, masts are pretty efficient beams, with a pinned bottom connection, tables of Ixx & Iyy for various extrusion sections.
3) While sailboat masts all have a strong, hinged bottom connection (a 'tabernacle') to the hull, there is a lot of variability in the strength & flexibility of decks.
4) Between the mast & deck, boaters may use shims, blocks, wedges, etc., of wood, hard rubber, cast acrylic, etc. To get one of these seated, it is pounded in while loading the mast in the opposite direction.
5) Some boats use a threaded 'tierod' to connect the mast to the deck.
6) The maximum mast stress is not at the bottom, but probably at the sail spar(s) or deck.
7) There was mention of 'tuning the mast' which apparently means getting it seated with the deck and having all the cable stays properly tensioned. This gives the mast some rigidity to prevent it from flopping around when the direction of the sails change.
8) Finally, my opinion is that masts do need a 'break-in' or 'tuning,' the reasons being that a) the mast/deck connections need to be adjusted and seated uniformly after some flexing, and b) the cable stays need to be properly tensioned. With too much 'play,' a sudden shift in wind direction could create a whiplash effect which would bend the mast. Both reasons allow for more uniform stresses on the mast.

Work hardening and age hardening are not involved.

 

[AlanD]

Unfortunately Kenvlach, those comments are not relevant in this case. The boats in question are the International Laser Class (one of the Olympic classes). Although people in the past have thought that work hardening is occurring, the people currently pushing the issue are the 5 manufacturers around the world who are trying to avoid replacing masts that are deforming far too easily. The real issue is that the specifications are not up to scratch.

My background is as a professional metallurgist not currently working in the industry, however I'm also the Australian Measurer for the class. Most people in the class will accept what the manufacturers say with regards to "breaking in" the mast, but as a metallurgist I cannot. The manufacturers are informing us that the masts should not be used in winds over 12 knots, until the masts are "broken in". A good quality mast can easily survive 30-40+ knots, which is beyond the skills of most people in the class.

The mast are a two piece arrangement. Both sections are circular anodized aluminum tube, with the bottom section being of a larger diameter than the top section. A plastic collar locates the bottom end of the top section within the top end of the bottom section (305mm/12" overlap). The fittings on the mast sections are all under compression (unless you do something stupid on the water).

In my opinion I believe that the real issue is that tolerance specified by the boat manufacturers are inadequate. The problem lies with the wall thickness, alloy or properties after aging.

I hope this background information assists.

My original problem still remains, does anyone believe that work hardening could be occurring below the yield strength. In my opinion it cannot occur.

 

[kenvlach]

AlanD,
I don’t fully understand the boats, and there may be signicant differences between Olympic class and ordinary sailboats that I drew conclusions from. However, I feel that you haven’t disproved my basic premise [the mast needs to be uniformly seated and the rigging ‘tuned’ in order that the mast transfer stress to the deck and hull without any excess play that could create localoverloads].
However, you have brought up some points for consideration. Also, I have read that some sailors use stainless steel cables while others use coated, high strength steel, and I presume that some people use polypropylene rope, and maybe racers use some Kevlar type.
So please enlighten us a bit more.
1) Are the masts similar to those shown by Dwyer

http://www.dwyermast.com/

or is there a geometric difference between ordinary & racing masts?
For example, do the latter omit the internal conduit for electrical?
2) At what location on the masts does failure typically occur (relative to hull, deck, splice, attachments and loading points)?
3) Do the masts fail in strong gusty or change-of-direction winds (rather than strong steady wind)?
4) What kind of rigging cables are used? At what tension? Maybe some work hardening?
5) What kind of mast-to-hull connection? What material? Does it need adjustment?
6) Do the mfrs. recommend break-in for shorter, one-piece masts?,
7) By “The fittings on the mast sections are all under compression” do you mean that through-bolts are used, with a piece of rigid tubing inside the mast to avoid collapsing it?
8) You say “the real issue is that tolerance specified by the boat manufacturers are inadequate. The problem lies with the wall thickness, alloy or properties after aging.”

Since you believe that no work hardening occurs via break-in nor that age hardening occurs after receipt of the mast, (and also disbelieve my idea that ‘break-in’ is getting rid of slack in the mast connections), you believe that ‘break-in’ is nonsense. Hence, some masts will fail at 16 knots due to mfr. defects regardless of break-in, and some better made ones will withstand 40 knots regardless of break-in, contrary to the mfr. spiel. Is this a correct statement of your opinion?

 

[AlanD]

Photos in the link below

http://www.mhasc.com/content/gallery/group1/index.html

Measurement diagrams here (note these do not specify diameter, wall thickness etc),

http://www.laserinternational.org/rules/measdiag.htm

1. Lasers are a one-design class; we can only purchase equipment from authorised dealers, who obtain that equipment from licensed builders. They are not tapered sections, they are just like the tube you could purchase from your local aluminum supplier, however the diameters/wall thickness etc are obtained by using class owned dies.

2. The usual point of bending and failure is near the collar 305mm position on the top section). The bottom section will usually bend near the 945mm fitting where the boom attaches, but fail near the 445mm fitting just (25-50mm) above the deck level. Obviously these are stress concentration points, but usually any failures are the result of corrosion.

3. The general use failures tend to be in stronger winds and often associated with the masts hitting the water i.e. capsizing. The bending of spars is a well known phoneme and would normally occur in stronger winds, but we are having problems with masts bending in light breezes when they are first used (IMO a quality/specification issue).

4. In the case of lasers, the masts are what are known as free standing, in other words there are no cables (stays) holding them up.

5. They are seating in a loose fitting fibreglass tube, which joins the deck to the hull, approximately 400mm long. These tubes are fixed and do not require adjustment.

6. They make the same recommendation for all mast/boom section. A small note, the sections used for the radial bottom section shown in the second link, is of a thinner wall construction which has a smaller diameter sleeve inserted, these sections are currently causing the most concern at the moment, but the problem still exists for all other sections.

7. The fittings are riveted onto the tube, but all rivets are located where the tube when bent (flex load, not yielding) would be under compression. By that I mean that the mast flexes backwards and the rivet holes are on the backside of the mast, there is also some sideways bend. In the 18 years I have been sailing this type of boat, I have bever seen a crumple failure, where the mast wall has collapsed.

8. Hence, some masts will fail at 16 knots due to mfr. defects regardless of break-in, and some better made ones will withstand 40 knots regardless of break-in, contrary to the mfr. spiel. Is this a correct statement of your opinion?

Yes (and the wind strength can be as low as 8 knots).

 

[BespinSunset]

I may be completely wrong with this but I'll give it a shot anyway.

The masts are constantly being loaded and unloaded. The manufacturers anticipate loads which may cause concentrated stresses beyond yield and subsequent workhardening. With the load removed the mast may obtain a slight permanent set which after first usage may be localised at part of the stress concentration.

As time goes on and the load is applied in different directions the permanent set would become more uniform around the stress concentrated area. Subsequent reloading would then result in a more linear relation between the strain and the force applied and workhardening would not continue until a load higher than the yield stress were reached. The unloading of the mast would also result in a return to a state that appears undeformed.

 

[CoryPad]

My opinion is that strain hardening cannot improve significantly the performance of these masts. The strength change is only ~ 10%, and that would be after significant plastic deformation. I don't see where the manufacturer can claim any improvement vs. time. Has there been evidence that "broken-in" masts still fracture at relatively low stress?

 

[Metalguy]

If alum. masts had to be "broken-in" metallurgically prior to "full loads", I would imagine that the same idea would apply to aircraft wings-but it doesn't.

 

[kenvlach]

AlanD,
Thanks for the details and photos. You've given us a much better idea of the situation.

Any close-up photos of bent masts in-situ (not removed from boat, with all fittings still attached)?

Re degree of plastic deformation: with the 6061-T6 numbers given by Corypad and an elastic modulus of 69 GPa, the yield strength is reached at a deformation of 0.40 % and the UTS at 0.449 %. Not much leeway. The work hardening would occur during while the tension side of the mast is stretching on its way to failure.

While corrosion is certainly a factor in long term failures, we should restrict this thread to the short term failure issue: Is the 'breaking-in' of masts scientifically sound or the mast mfr.'s hocus-pocus to cover-up material problems?

 

[AlanD]

BespinSunset: I know where your coming from and in my opinion it's basically the only way you could achieve work hardening, except that the work hardening is occurring during the natural flexing of the mast section. So when we come off the water and remove the masts, no permanent set may be observed under normal circumstances. Class rules prevent us from sailing with no straight mast sections, so we have become adept at straightening them.

Corypad: No evidence of "broken in" masts failing at low stresses, except through corrosion problems later in their life.

Kenvlach: Sorry no photos are available of bent mast sections in situ. I'll see if anyone bends one this weekend where I normally sail.

"Is the 'breaking-in' of masts scientifically sound or the mast mfr.'s hocus-pocus to cover-up material problems?"

Hence why I asked this question here.

 

[TVP]

"Breaking-in" of the mast assembly may be happening, but strain hardening of the mast due to elastic flexing is not. I doubt that strain hardening of the mast due to plastic deformation is happening either. The scientific principles of materials science explain why strain hardening does not occur under elastic conditions. It sounds like mast manufacturers need to investigate:

a) stresses on the mast during operation, including a detailed analysis based on the exact degree of fixity and constraint imposed by the connections

b) statistical distribution of incoming mechanical properties (mostly yield strength)

c) potential variation in heat treating of alloy 6063, which by the way, is a great alloy for extrusion, but not necessarily a great alloy for end users. Response to heat treating is questionable in my opinion.

 

[BespinSunset]

AlanD

It's a good query!

It is the only way I can imagine work hardening specifically to be of benefit in this case.

 

[brad]

Just a few thoughts that nobody has mentioned above:
As I understand this, these boats are designed at the edge (without much of a factor of safety). I make this presumption because speed (hence mass) is enough of a factor that one would not any more mass than is necessary. Please correct me if I am wrong on this.

Given this assumption, one would presume that this is designed to operate at or near "yield point". Keep in mind that classic "yield stress" is at 0.2% plastic strain. All materials diverge from Young's modulus (i.e. "go plastic&quot at a point lower than the book "yield stress". I would presume that these masts are designed to operate near yield strength, hence there may be some amount of work hardening. I make this statement not as a sailor nor nautical engineer; I am building up from the problem as I understand along with an understanding of mechanics.

I would argue that work hardening CAN happen below "yield". I cannot state how significant this could be.

Note--MetalGuy conjectured that this would be expected to also hold for aircraft wings (therefore proof that this is not the case). I contend that this is an apples-to-oranges comparison, as the factor of safety on aircraft wings is certainly much greater than the factor of safety for this application (hence the operating stresses are MUCH lower for the aircraft).

Brad

 

[AlanD]

Interesting point Bradh, which I really can't comment on, I lack the knowledge of the original designers. The boats were initially designed as a "fun" boat, rather than a "racing" boat. The Olympic status was gained 25 years after the boats were first designed, but with the exception of tightening up some of the tolerances of the fittings (+/-12mm reduced to +/-5mm). My assumption would be that the original designers selected the alloy, mast diameter and wall thickness on what wouldn't bend in ordinary usage as a fun boat.

 

[kenvlach]

There have been many good comments, but none directed at the actual amounts of bowing of the masts in the photos and dimensions in the 2 links provided by Alan. Maybe we need more understanding of the particulars; at least, we non-sailors do.

Alan,
I was rather amazed by the dimension schematics; straight mast sections, but the lower figures with sails show considerable bending (unfortunately, not to scale). Observations and questions:
a) Is there a Class numerical value for max. bending allowed in a strong wind? Or
b) Is there a max. allowed ‘preload’ on the mast that bends it backwards (from tightening the really strong sail (maybe with tension cables sewn into the sail) connecting the boom & mast)?
c) The sailor for a given wind) controls the amount of bowing with the boom; with loosed boom and almost slack sail (photo marapov, lower left with green buoy) the mast is straight, while with a taut, bowed boom with a strong wind (photo 11agstrt, just left of center) the mast shows incredible bowing. So, I presume the boom has a swivel attachment to the mast there is a lot of furious cranking on some winch during racing. Correct?
d) Last but not least, with all tension is removed, does the mast return to its initial straightness?

Suggestions.
As a beam model for loading of the mast, I suggest that the mast, boom and sail act as a cable-stayed jib crane. The tensioned stays (integral with sail) from the end of the boom radiating to different sections of the mast account for the bowing because the vertical component increases with the sine of the boom-stay angle. This keeps the rear of the mast in compression despite a rearward, local tension upon the mast due to wind in the sails (opposite of a gravity load upon a jib crane).

Maybe Alan can confirm this, give us the mast ODs and someone with a suitable computer program can input some strains and ‘back out’ a reasonable wall thickness for the mast and a presumed stress at the elastic limit.

 

[Metalguy]

Unless these masts are guyed in some way such that they are under constant bending loads, in all probability they are designed primarily to prevent fatigue cracking. If this is true, they would not be loaded anywhere near their proportional limit or yield strength. The designer must decide the approx. number of cycles and stressing conditions the mast should withstand.

Given the orders of mag. more loading cycles on aircraft wings during the expected life of an airplane, the comparison would then be valid.