why 2017 AND 2024 RIVETS are always T4 Configured,..? 6

Rivets have to be ductile enough to be upset properly at installation.
If you let them get hard before installation, they'll fracture instead of plastically deforming.



Mike Halloran
Pembroke Pines, FL, USA

To answer the other part of question--refrigeration is necessary, as solution treated alloys (T4) will start to naturally age at room temperature and become harder/stronger.

refrigeration applies to DD rivets, D, and AD rivets don't age at room temperature.

From AR-MMPDS-01
T3 solution heat-treated*, cold worked, and naturally
aged to a substantially stable condition.
Applies to products which are cold worked to
improve strength after solution heat-treatment, or
in which the effect of cold work in flattening or
straightening is recognized in mechanical
property limits.
T4 solution heat-treated* and naturally aged to a
substantially stable condition. Applies to
products which are not cold worked after solution
heat-treatment, or in which the effect of cold
work in flattening or straightening may not be
recognized in mechanical property limits.
* Solution heat treatment is achieved by heating cast or wrought products to a suitable temperature, holding at that temperature
long enough to allow constituents to enter into solid solution and cooling rapidly enough to hold the constituents in solution.
Some 6000 series alloys attain the same specified mechanical properties whether furnace solution heat-treated or cooled from an
elevated temperature shaping process at a rate rapid enough to hold constituents in solution. In such cases the temper
designations T3, T4, T6, T7, T8, and T9 are used to apply to either process and are appropriate designations.

so T4 rivet blanks are driven into a T3 condition.

The OP asked about 2017 (D) and 2024 (DD) rivets. Both of these are usually considered Ice Box rivets. 2117 (AD) rivets are driven at room temperature. All three alloys are a T4 temper.

caught !

Engineerpaki... to elaborate on MikeHalloran's statement.

2017 [D] and 2117 [AD] rivets can be driven "cold" in the –T4 condition: there is adequate ductility to prevent cracking and maintain toughness. HOWEVER… the D material has a significantly higher yield strength, so it requires MORE energy to deform this alloy than the AD… leading to a practical limit for hand driving.

The 2024-T4 [DD] rivet requires even higher driving forces and cold-works rapidly: for all practical purposes, this rivet alloy cannot be hand or machine-driven “cold”. In every case these rivets are have to be solution heat treated and quenched to attain the after quenching [AQ] ”–W” temper. The –W temper is highly unstable in that it will spontaneously start hardening to the –T42 temper: however it’s prime advantage is that the –W temper material becomes temporarily very ductile and attains a very low-yield strength [~8-KSI FTY]. In this condition even large fasteners can be driven successfully by hand; or by machine [squeeze]. Once set-in-place by cold-working, it spontaneously ages [hardens] to the –T42 temper.

NOTE: DD rivets are heat treated to the-T4 temper at the factory, so they will NOT be susceptible to dents/dings, distortion, etc during handling… and POSSIBLY to accept the anodize finish better [dead-soft annealed aluminum alloys are difficult to deal with].

NOTE. 2xxx alloys may be frozen [below 0-F] within a few minutes after quenching (just enough time to dry-off the quenchant), which will stop (almost) the spontaneous aging process (sorta puts the alloy-transformation process into hibernation). This allows technicians to transport the rivets from the HT shop to the work site in a freezer, and install the “soft” the rivets in small batches over a very long period [not the insanely short period if allowed to warm-up]. The process, and these rivets, are called “ice-box rivets”… so-named for the refrigerated transport container “ice box” [often an insulated container with dri-ice @-100F.

NOTE. Alloy 2017 rivets can also be heat treated/quenched/installed using this same “ice-box process”, making them considerably easier to install in large diameters. NAS523 has (2) different codes for the D rivet, depending on the install process.

CAUTION/WARNING. The “–W” temper is NOT the same as the annealed “-O” temper, which is a long-term stable, low-yield temper [created by a different HT process].

CAUTION/WARNING. Rivets intended for solution HT and quench by the shop MUST have bare anodized finishes. Chemical conversion coatings will be “burned-off” during the SHT phase. Sealant and primer will stick very poorly to what is left of the finish [ash?]... and they will be vulnerable to corrosion. Color-dyed anodized-rivets will fare a little better: the dye-color [red, orange, etc] will be burned-off the surface, leaving a poorly adherent finish.

2xxx-alloy -T62 [user HT] or the -T81 [OEM] tempers, are developed by a post quench moderate heat follow-on bake for ~12+hrs. This temper raised the FTU slightly… but dramatically raises the FTY significantly closer to the FTU. At this low ductility and high yield state, cold-driven rivets WILL crack or possibly shatter/spall in the bucked tail.

Now… lets get practical… regardless of whether AD, D or DD –T3 [-T42] rivets are installed cold or “ice-Box”, the rivets are NOT going to over-age to –T62 or –T81 without high heat over an extended period to “push” to process along. By definition, the heat/time required for this temper transformation would grossly affect the temper of the structural alloys thru which the rivets have been installed.

Trivia NOTE. “There is evidence” [poorly documented for obviously reasons] that 2xxx–T3/-T42 alloys can “migrate temper” towards the –T62/-T81 temper range if left in a high heat environment for decades… such as a desert climate


Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Ice box rivets harden significantly with age. To prevent age-hardening of the annealed rivets prior to installation, they're stored cold.

Is annealed the correct term to use in this case? Aluminum is solution heat treated, which means to heat it hot enough to where all the alloying elements fully dissolve. This is below the melting point. When quenched rapidly the alloying elements do not have time to form crystals or grains so the aluminum is soft (low yield strength). Steel, on the other hand, is hardened by quenching. At room temperature the grains grow with time, which increases the strength of the aluminum (natural aging), so this soft state is not stable, but can be retained by freezing. Heating aluminum to about 300-350F allows the grains to grow in a period of hours to an optimum size for maximum strength (artificial aging). Heating too long or too hot allows grains to grow very large where they do not strengthen the aluminum and the aluminum becomes soft again. I think that this over-aged state would more properly be called annealed. Maybe I'm wrong, but calling both soft states "annealed" is ambiguous and confusing.

I've got some AD rivets that were passed down to me from my dad and are WWII vintage. It's supprising how the 60+ years has allowed them to age harden compared to recently produced rivets of the same size and alloy (AD).

Compositepro... lots of folks get HT concepts WRONG.

NOTE. See Aluminum Association spec AA H31.1 and/or SAE AS1900 for a basic discussion of aluminum alloy tempers.

Annealing is a process where a metallurgical structure [aluminum, ferrous, nickel, titanium, copper, etc alloys] is taken to its most ductile and hence most deformable, but long-term-stable, state. For most metals annealing is accomplished by heating the alloy to a specified temperature for a specified soak time... Then cooling SLOWLY... back to room temp. Aluminum alloys are typically cooled in still air on a rack; steel and other alloys highly prone to oxidation may be cooled in the oven with inert gas to eliminate oxygen, etc]. Most alloys have various stable annealed conditions. For instance 2XXX and 7XXX aluminum alloys can be conventionally to the -O ["Ooo" NOT -0 "zero"] temper... or the specially annealed state called -O1 ["Ooo one"] temper where grain structure is slightly harder/tougher. NOTE: the temperatures for annealing process are typically close-to to solution heat treating temperatures... but not quite the same.

Solution Heat Treating is a process where a metallurgical structure [aluminum, ferrous, nickel, titanium, copper, etc alloys] is taken to certain grain state at that is only stable at the elevated temperature. The alloy is then RAPIDLY cooled [quenched] to room temperature... typically within a matter of seconds... the high temperature grain structure is frozen down to room temperature [below 140F]. This rapid cooling can be accomplished by agitated immersion in water [with mixtures to delay boiling to a higher temperature and minimize in-water oxidation], oil, cold gas, etc. The resultant metallurgy is then either in a an unstable transition state... or in a stable temper state where further transition is feasible/necessary.

For 2XXX, 6XXX and 7XXX heat treatable aluminum alloys the post SHT After-Quench ["AQ"] temper is the "-W" temper. At this point aluminum alloys ACT like they are annealed... for a short period of time... however, there are a LOT of crazy metallurgical process occurring within the alloy. For 2xxx alloys there is a spontaneous transition to the -T42 temper that occurs within ~96-hours. This transition rate-of-reaction can be delayed [freezing] or accelerated [by heating]... typical chemistry 101 stuff. I think the copper within the 2XXX alloys is the driving factor here. NOTE: 6xxx and 7XXX alloys typically have to be pushed by heat from the -W temper to a more useful strength temper.

Interesting note. When an 2xxx or 6xxx aluminum alloy is deformed in the -O temper "the-grain-remembers-the-strain" and the grain typically grows proportional to the strain. When SHT/Quenched/Tempered [heat-aged-hardened] the grain grow is still evident in the heat treated part [-TX] and MAY be detrimental to overall strength/toughness. Hence the -T42 and -T62 tempers were developed to describe/quantify these less than ideal temper states.

However, when alloys are strained in the unstable -W temper [within a certain critical period of time, AQ], then "the-grain-forgives-the-strain" due to the unstable metallurgy. EVERY MATERIAL SPEC [AMS] for heat treated raw sheet/plate, extruded shapes or drawn bar have a basic temper [-T3, -T6, -T7, -T8] allows for the beneficial effects of strain that occurred AQ [in the -W temper state]... and the last digits of the temper number -Txx50, -Txx51 -Txx510, -Txx511, etc are added indicating the type of strain, AQ etc.

UH OH, my brain hurts... this quick reply has gotten more complicated than I expected... gotta go back to work.

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Compositepro said:

Is annealed the correct term to use in this case?

Click to expand...

Sorry about that, you are correct - definitely the wrong term on my part. Solution heat treated is correct.

I've been working with nickel stuff for too long (annealed = soft). What I was getting at was "the rivets must be kept cold because they otherwise age harden."