I have always wondered why magnesium rivets were used in British aeroplanes during the war. Magnesium acts as a sacrificial anode to aluminium, resulting in the phlegmatic restorer of today picking magnesium rivet heads off with their fingernail. Over time, electrolytic corrosion causes the magnesium to shed electrons to the surrounding aluminium, effectively dissolving the rivet away. I have heard stories of dismayed buyers of original Spitfires picking rivets out of wings. The original constructors knew all about electrolytic corrosion – why use magnesium rivets then ?
The obvious answer is that these aero structures were not meant to last. They were designed for a life of hours, not decades, and not to please unborn millionaires. But surely there must have been a more practical reason. Aluminium was a strategic material, not abundant, but the gross weight of rivets in a structure could be measured in pounds, and I cannot believe a marginal weight saving could drive the introduction of a novel material into the production process. Magnesium too, was a strategic material, not abundant, most advantageously used to lighten the great mass of aeroplane wheels.
I have had a suspicion that the use of magnesium rivets was driven by the advantage of avoiding the need to heat treat Duralumin rivets. Duralumin was the aluminium-copper alloy that developed great strength after heat treatment and ‘age hardening’. When heat treated it became soft, then progressively harder over hours and days. In production, this would have created a significant task in preparing heat treated rivets to align with riveting work scheduled for half a day later, then a great urgency to drive rivets while the material was soft. The age hardening process could be slowed down by refrigerating the Duralumin, but this then introduced the additional task of refrigeration, whether through the use of dry ice or freezer rooms in a plant.
So here was a complicated chain of production steps, where any disruption to one step would frustrate the rest, grievously affecting total production. Here was a need to establish dedicated heat treatment and refrigeration facilities for Duralumin rivets and, almost like the powder monkeys of Wellington’s battleships, rivet monkeys rushing trolleys of age hardening rivets through plants, prodded by anxious foremen gazing at pocket watches. God ‘elp if a falling bomb disrupted gas supplies and put the furnace out. There had to be a better way.
So here was the magnesium rivet, not quite as strong as the Duralumin rivet, but able to be work hardened. By simply hammering the rivet head, the rivet might develop the same strength as a Duralumin rivet. This thought first came to me when discussing the working of magnesium sheet through an English wheel. Magnesium sheet was good for two passes and then it ‘work hardened’ and became reluctant to form further. I was more familiar with ‘work hardening’ in stainless steel sheet; magnesium was a little more exotic, more familiar in German aero structures.
So here the flower opened up : magnesium rivets could eliminate the need for heat treatment and refrigeration, free up staff, eliminate a great vulnerability in production scheduling, speed production. It was irrelevant if, in a number of years, the rivets would fail, as electrolytic corrosion took hold. Nazi bullets would no doubt account for the structures sooner than that. I sense this was an entirely British act of pragmatism. I have not heard of magnesium rivets being picked off old Kittyhawks or Corsairs or Dakotas. This habit may have crept into perhaps Australian production work. It seems to be something that happened when the devil was at the gates. I do not know, if in the more serene production plants of California, Duralumin rivets continued to be used exclusively, to the relief of modern operators of B-29s and North American Mustangs.
These are musings which had rolled around in the mind for some time. I would be grateful if anybody could confirm or deny or add to this. Today, with some time on my hands, I delved into the old books to see if I could put some engineering meat on the bone. By cross referencing a list of wartime DTD Standards, British Standards and the most excellent Nov 1944 RAAF Aeronautical Engineering Handbook, I propose the following :
1. Two aluminium rivets types were used in wartime British construction, BS L36 (pure Aluminium, 15 T tensile strength) & BS L37 ( Duralumin, heat treated, 56 T tensile strength) Tensile strength is an appropriate dataset for rivets, being effectively the resistance of the head of the rivet to failure. For high performance aero structures, L37 – 56T rivets would be used.
2. The most appropriate wartime magnesium material for the making of rivets was DTD 259 (38 T tensile strength). This tensile strength could be increased by work hardening, effectively the forming of the rivet head. Would work hardening develop an additional 18 T of tensile strength, to equal heat treated BS L37 aluminium rivets at 56 T strength?
A confirmation of this, or alternate theory that could make me pause between mouthfuls of pint with pursed, thoughtful lips, will result in a pint for the winner, unless you are in the Falkland Islands, in which case some sticky paper to stop penguins from walking over your daffodils will be your postal prize.
By: Dev One - 18th April 2016 at 13:07
Flat head rivets, I must admit that in all my years in the industry, making &/or designing, I have never seen or specified one!
Badly driven rivets….have seen some horrors there; I remember seeing the first Viscount wing that came to Weybridge for re-sparring, double holes & bent rivets (nailed over, not formed correctly), it was apparently the first wing produced at Hurn.
As for riveting from the outside, yes done a few, usually because of lack of space on the inside to get a gun. Usually the head is formed using a domed headed ‘set’ on the gun.
Keith
By: Robbiesmurf - 17th April 2016 at 20:57
AS2229/2230 countersink are 90 or 120 degrees, indeed dependent on the material thickness. That has to do with the area of influence?
The SBAC standards be it AS (Air Standards) or SP (Standard parts) are also supplied with a few different proud heads. There is Snap, which is a small diameter, high head, Mushroom, which has a larger diameter and lower height and of course, Flathead, which speaks for itself. The MS series are as you say countersunk to 100 degrees (MS20426). The proud version (MS20470) is actually called Universal head. There are also repair rivets with a NAS (National Aerospace System) number. They have a larger shank diameter to the head diameter. They also come in various strengths (and material specs) , A, AD, D, DD & B. They are identifed by various markings on the heads. A=plain AD= sunken dot D= raised dot DD= 2 raised dashes B= raised cross. An interesting side note, the B series are used to join magnesium alloys because of their corrosion resistant properties. The B’s can be stored indefinately without any change in it’s hardness. I assume they are a magnesium alloy.
Hand-built rivet cutters were part of the apprenticeship course if I remember correctly, along with rivet squeezers.
Your point about the different c/s angle on the Valiant. I remember a Douglas a/c having a skin repair in the UK, many years ago. The skin was countersunk to some 87 degrees and the rivets were driven from inside to out (universal head). After that they were all milled on the outside. Quite a number of rivets had to be re-done as they had not been driven correctly.
By: Dev One - 17th April 2016 at 20:00
From memory, AS countersunk rivets had 120° heads, & I think the Mush heads were of a different profile to the SP series. There were also 70° csk heads that were used on the Valiant, but they were heat treated ( as an apprentice one of my jobs was to reload the rivet container from the heat treatment department every 2 hours!) If they were more than 2 hours old the tails would split open. These 70° heads were riveted proud so that some poor chap afterwards had a compressed air router to mill the heads flush with the skin.
MS rivets have a different head profile apart from the 100° variety; Brazier & Pan head come to mind (which is fading fast!)
120° were generally used in dimpled holes in thin 22 swg or thinner skins, Viscount & earlier had these.
Most of my riveting was done with compressed air hammers that came in various sizes to cater for small or large rivets, (2X, to 6X sizes I seem to remember).
There were also hydraulic ‘scissor’ type riveters as well as manual ones to squeeze the rivet. Squeezing being restricted of course for accessibility reasons, but was preferred design wise as the finished product was considered stronger. There were of course those for snap heads as well as csk one side.
One of the other tricks was to cut the rivet to the correct length, as back then there did not seem to the variety of lengths available.
Keith
By: Robbiesmurf - 17th April 2016 at 16:05
Most of the ally alloy rivets do work-harden when rivetting. I have a selection of AS and SP types running the range of materials being L36, L37, L58 & L86 (black, natural, green & violet), AS has been superceded by SP standards and are I believe interchangeable. I also have bucket of MS-AD rivets that have hardened already. One day, when I get a proper oven….. I have used a number of the AGS (British) rivets for a couple of small projects and have experienced the work-hardening. The trick when rivetting is buzz them in one go, maybe with a slight touch afterwards. Continuous hammering causes them to become more resistant….
What kind of equipment do you use to rivet?
By: powerandpassion - 17th April 2016 at 14:07
Dural – Duralumin- Duraluminium was a proprietory description that became a catch all for a number of related alloys, all heat treated. the HT alloys always developed more strength and more dependable strength that cold worked materials, so I understand they were always used in spars and similar structural members.
I would say the HT alloys, while in the soft condition, were easier to work than cold worked materials. Magnesium would become harder and harder as it was worked. Most of the old Magnesium rivets I have seen seem to have got one single, good pounding to form the head.
DTD327 was still a Dural type alloy, in looking through the books this seems to be a RR alloy, RR75, with an exotic touch of Titanium and Nickel, so probably quite a performance improvement, but still requiring heat treatment to develop workability, then strength through age hardening.
In terms of DTD 303 becoming BS L58, this British Standard is given as ‘Cold forged rivets of Aluminium 5% Magnesium alloy’, which I interpret as an Aluminium with 5% Magnesium. It will be interesting to compare this with the original DTD 303 specification, if it can ever be found. Often I find the materials in Standards change even though a later standard cites the original DTD document as its source, perhaps driven by a performance aspect such as ‘cold forging’. For the restorer looking for material substitutions, this is can be an unreliable connection.
I would dearly love to find an old trunk full of original DTD specifications.
Please Santa.
By: Dev One - 17th April 2016 at 13:11
I notice also that Dural rivets are still called up through the spar web on the bit of the drawing shown, so I would agree with Robbiesmurf re the production improvement, because the DTD303 material would need a lot less force to peine over & could possibly use hand squeezer type tools rather than hydraulic or compressed air driven guns?
Keith
By: Robbiesmurf - 17th April 2016 at 11:48
I’ve just read up on the rivets (AP101A-1401-1 Dec 85 amdt11). The DTD.303 standard was changed to BS.L158. It’s main carachteristic is that it doesn’t require heat treatment. Although being slightly weaker, I suspect the rivets were introduced to speed up production. A service-life of 20 years + was not really expected at the time… Another factor could have been the available materials, the HE162 is a good example of that.
By: powerandpassion - 17th April 2016 at 05:56
Spitfire
Attached pics courtesy of Airframe Assemblies show Spitfire rear spar drawing, first issued in 1937 showing the rivet requirements. Pictorially still showing Dural L37, but with a revision note dated 1942 changing things to DTD 303 (Magnesium) or DTD 327 (RR75, an aluminium alloy).
[ATTACH=CONFIG]245391[/ATTACH]
[ATTACH=CONFIG]245392[/ATTACH]
So here you have official confirmation of the use of magnesium rivets in wartime aluminium aero structures. Chris from AA maintains that in his experience of original Bf.109e structure there was no sign of magnesium rivets, or magnesium sheet material for that matter. Magnesium was a material most familiar to German constructors, yet the hard headed decision to use magnesium rivets seemed to only have been taken by the British in 1942. I wonder if the Germans had switched over by 1944, when things were dire for them. Next time I am near an original ME 262 I will have to take out the car keys and see if I can pick off some rivets…
I can confirm that magnesium rivets were not used in Australian manufacturing, for what ended up being the fourth largest Air Force in the world in 1945.
The use of magnesium rivets did, in fact, provide the electrolytic corrosion protection mooted in previous discussions : the rivets corroded in preference to the aluminium sheets, protecting the larger aero structure. This accident could inform an adaptation of the concept using a single, easily accessible magnesium disc attached to the structure, that could be replaced every 5 years.
By: Robbiesmurf - 13th April 2016 at 07:36
Well, we do have a product called PX1, a lanolin suspension. When used correctly, it can protect metals for decades.
By: John Green - 12th April 2016 at 20:36
There is another line of defence against electrically induced corrosion. The incorporation of a galvanic isolator will prevent the establishment of a potential circuit that has occured due to the presence of differing metals in an electrolyte medium.
Galvanic isolators are commonly fitted in boats because of the potential for unknown electrical activity and therefore corrosion in moored vessels in marinas. I’ve seen anode protected sail drive units ( the bit that turns the prop) experiencing moderate to severe corrosion within one year of being in the marina. Perhaps historic aircraft could similarly benefit from such an item.
By: Robbiesmurf - 12th April 2016 at 13:50
Indeed, sacrificial coatings and parts is not new and has many applications. The Sidney Opera House uses zinc blocks to protect the steel pipes of the air conditioning which are cooled by sea water. Zinc annodes are also used in simple water boilers in homes these days. The rivets I used are both of the same dimensions (0410= 1/8 dia 5/8 long), heads are both 90 degree and the dimensions are the same. That is why I chose those. I think the choice of material has mainly been down to availability at the time. Post war Landrovers were built out of large amounts of re-cycled Spitfires. Steel was short at the time, it was spread all over Europe in 25-50 ton blocks at the time. Even the pop rivets used on the Landrovers came from the AGS stock…
By: powerandpassion - 12th April 2016 at 13:03
I happen to have a few hidminiums (AS2229 0410) and loads of the magnesiums (AS162 0410). I weighed them with my scales and came up with this:
70x AS2229 0410 = 26 grams
70x AS162 0410 = 24 grams
Robbiesmurf,
Thank you for doing this little weight test.
Given that magnesium is lighter than aluminium (really aluminium copper alloy) then there must be more metal in the magnesium rivet to come up with an almost similar weight. So the other way to develop more strength in the magnesium rivet, understanding the shank must be the same diameter, is to put more metal in the head. So you could say ‘more metal & work hardening’ is the principle behind interchangeability with duraluminium rivets.
Courtesy of a few little chats on the subject this seems to be an entirely UK issue, eg Beauforts made in UK had magnesium rivets but Beauforts made in Australia had aluminium rivets. I do not thing that ‘designed obsolesence’ was the factor behind this, rather, faster production. Also distributed production, without the need for heat treatment facilities, explaining why final assembly works incorporated some structural sections with magnesium rivets.
I do not think aircraft disintegrating in mid air due to electrolytic corrosion of magnesium rivets was ever an issue in wartime, but certainly after 10 years you would start to be nervous. I wonder how much the reluctance of postwar aviation regulators, staffed with people that truly understood the structures, to support the nascent warbird movement in the 70’s developed from this issue ?
Turning a negative into a positive, I do believe that historical aeroplane structures should be fitted with sacrificial magnesium anodes, in the same way that this practice is readily accepted in the marine world. Today, we do want aero structures to last 100 years. So a readily accessible and replaceable lump of magnesium will protect an aluminium structure. It’s the same as zinc based paint, except magnesium will more actively protect the aluminium. For a few quid, a Museum can protect static airframes exposed to the weather. For aluminium parts sitting on a shelf, I use a band of magnesium ribbon, purchased on ebay, to monitor and protect. I have seen too many steel shipping containers holding aluminium parts, subject to daily cycles of condensation, where the aluminium sacrifices to the steel. A part ‘protected’ in goodwill by a volunteer putting it in a shipping container will be swept out as dust by another volunteer in ten years time.
By: powerandpassion - 12th April 2016 at 12:31
the powder monkeys of Wellington’s battleships,
Apology and Retraction.
It came to me at night that the Duke of Wellington was an Army man, and that what I should have written was ‘Nelson’s battleships’. I apologize if I have offended anybody in the Nelson family, much as if I had said ‘Montgomery’s Bomber Command.’ Still they don’t teach us colonials this stuff anymore and none of you lot picked up on it either, so fat lot of good we all are in a pub quiz!
God Save the Queen !
By: Robbiesmurf - 11th April 2016 at 21:29
Avoid the MS rivets then. They do age harden and be a right little……….tinker when using. To soften them you’d have to heat treat them. 450 deg C for about 30 mins, after quenching, keep them in the freezer until needed. The treatment normally lasts about 20 mins. Freezing them slows that down.
By: Arabella-Cox - 11th April 2016 at 21:11
Thanks for the ID on the rivets. Actually most of them have numbers stamped on the tops rather than letters? They came in an empty coffee tin so there are thaaaaasands of them.
I was actually trying to find a cheap rivet set as I need to make up a few internal bits and pieces for the Horsa that are snap riveted. As their bit that are “on display” I can’t get away with pop-rivets. 🙁
By: Robbiesmurf - 11th April 2016 at 19:50
@ Powerandpassion.
Interesting points you made about the differences between the hidminium and magnesium rivets.
I got to thinking about the weight saving and put my OCD head on and had a look.
I happen to have a few hidminiums (AS2229 0410) and loads of the magnesiums (AS162 0410). I weighed them with my scales and came up with this:
70x AS2229 0410 = 26 grams
70x AS162 0410 = 24 grams
Both are c/s 90 deg rivets. It might not seem a great difference but rivet distances are in general 3-5x D (diameter), so the number nailed into an airframe is quite a number. I wonder what the weight saving would be. Have you counted the rivet holes?
When demanding them in the RAF, the D of Q was per 1/4 lb. The thing there being that if you had a bag of each, they would both weigh the same..;)
By: Arabella-Cox - 11th April 2016 at 19:16
Not a flier, but a static airframe. Anson 19 (mod), G-AGPG’s cockpit section had a nose extension, made in the late 60’s, to carry the bulkhead used to mount the weather radar scanner and attendant radome.
Fabricated from the usual aircraft materials the rivets were round-heads which, on cleaning in preparation for paint, began to pop off to the extent that it was likely the whole assembly would de-construct itself. Obviously not a good scenario; the “popped” ones were replaced and the remaining originals, eager to lose their heads, treated very carefully and re-finished. Virtually the only thing holding them on now is the layers of paint though, like I said earlier, not really a concern as it’ll never fly again.
There must have been an (written or unwritten) expectation at the time all this stuff was made that anything built in this way would only be around for a few years. It’s definitely a case of built-in obsolescence – or airframe life, so all who designed, specified or carried out such work would have been fully aware of the component’s limited life span. Presumably there were those same folk who knew this was the case and would plan accordingly. Whatever the ambient conditions in which the aircraft so constructed would be operated, it was a matter of certainty that, sooner or later the thing would, quite literally, fall apart.
Are there (m)any instances of aircraft falling apart due to rivet failure in flight?
Pic attached of Aggie Paggie’s (the Anson in question’s) nose area with its “popped” rivets.[ATTACH=CONFIG]245247[/ATTACH]
Anon.
By: Robbiesmurf - 11th April 2016 at 15:41
@ OneEighthBit.
Simple id of the rivet materials=
Black L36 tensile strength was 7 tons/sq in. Mark on head/tail A
Plain alu L37 tensile strength was 25 tons/sq in. Mark on head/tail D
Green L58 tensile strength 16 tons/sq in. Mark on head/tail X
Violet was L86 tensile strength 17 tons/sq in. Mark on head/tail S
You can start sorting them out..:highly_amused:
By: TonyT - 11th April 2016 at 10:33
The Wessex air intake for use in dusty climes was assembled with magnesium rivets, I assume to keep the weight down. When operated in the heat of a Sharjah day and a very cold night they started to fall apart. Hiduminium rivets were not affected.
As were a lot of the skins originally, if memory serves me correctly.
By: TonyT - 11th April 2016 at 10:32
OneEighthBit, now all you need to learn is how to knock those nails home, here you go
Sign In
April 10, 2016 at 12:21 pm