By Conrad Laskowski

Shape-memory alloys are specially formulated combinations of metals that are deformable when cold and return to a pre-deformed or “set” shape when heated above a transition temperature. Many different combinations of metals or alloys using aluminum, copper, nickel, gold, iron, titanium and/or zinc are now known to have shape-memory properties. Two of the most utilized shape-memory alloys today are copper-aluminum-nickel and nickel-titanium, the latter of which we will be “cooking” with here.

NiTi alloy is often referred to simply as “Nitinol” because of its origin in the 1960’s at the US Naval Ordnance Laboratory. The name “Nitinol” is a blend of the periodic symbols for nickel-titanium (NiTi), suffixed with the acronym for Naval Ordnance Laboratory (NOL). The shape-memory properties of materials like Nitinol make them possible alternatives to pneumatic or hydraulic mechanical actuators in some applications. Additionally, they can be activated passively by changes in system and environmental conditions, or actively by heating the elements, often done electrically.[1]

Nitinol – and all shape-memory alloys – change their shape because of transitions in the crystalline structure of the alloy in response to temperature and stress. It is true that other materials, like carbon-steel for example, will also change crystalline structure in response to temperature. Carbon-steel will go from austenite to martensite when heated and quenched. So how is this different from shape-memory alloys? In the case of carbon-steel, the changes are not fully reversible, whereas in shape-memory alloys like Nitinol, the changes in crystalline structure are fully reversible. It is the alloy’s ability to “remember” its prior shape at that crystal structure that gives it the unique quality of “shape-memory”.

So how does one get the shape memory alloy part they need?  Unless you are lucky enough to find a pre-set shape, such as a spring, available off-the-shelf from a vendor, fabricating a custom Nitinol part and then successfully heat-setting it to that shape can be tricky to do on a budget. Luckily, Root3 Labs has a hack.  First, the raw Nitinol material can be shaped and formed using most traditional metal working tools. Great!  Done (see Fig. 1).

Fig. 1 – The raw NiTi piece formed into the desired geometry.

Secondly, a fixture should be fabricated to hold the formed Nitinol geometry in place so it can’t reset to the stock shape (i.e. straight wire, flat sheet stock, etc.) when being heat-set. Be sure to make the fixture out of materials that can also withstand high temperatures. Perfect! Accomplished with the same metalworking tools and some scrap steel (See Fig. 2).

Fig. 2 – The raw NiTi piece fixtured in a thick steel jig.


Lastly, heat-set the formed shape so that it becomes the newly “remembered” default geometry.  Hmmm?  Easier said than done. There are several types of precise, high-temperature, uniformly heated ovens on the market for heat-setting shape memory alloy parts, but they run in the thousands to tens-of-thousands of dollars. Yikes. What about a blow torch (Christina Krueger’s favorite shop tool)?  Root3 Labs has found that using a blow torch until the Nitinol visibly glows red hot – as sometimes suggested – leads to overheating, poor performance, and localized failure of the part at “hot spots” through repeated cycles. So, what if you only need to generate a handful of Nitinol parts for a low-cost proof-of-concept? As is often the case, BBQ to the rescue! Or more specifically, smoldering charcoals to the rescue (Fig. 3)! Let me explain.

Fig. 3 – The (hot) dog days of Summer.


Instead of thousands or tens-of-thousands of dollars in a new precision oven, a $13.00 bag of charcoal can give you a “controlled enough” temperature condition to get your parts heat-set and working. Chef – *cough cough*… Senior Mechanical Engineer – Conrad at Root3 Labs selected a Nitinol material with an activation temperature of 176F (80C) and – more importantly – a heat-setting temperature range of 1,022F-1,382F (550C-750C). Typical peak temperatures for smoldering charcoals are in the range from 842F-1,292F (450C-700C), although very energetic and dense fuels such as coal can reach peaks at around 1,832F (1000C).[2]  It is important to recognize the difference between actively burning charcoals – where those peak temperatures ranging from 1,000C, upwards of 1,500C are often found – and smoldering charcoals where the more desirable temperature range for our purposes is just right. If you note that the high end of the smoldering charcoals’ potential is unlikely to exceed our max heat-setting temperature for the Nitinol, then we know that the worst-case scenario is that we may undercook the Nitinol. Therefore, the goal should be to surround the Nitinol and fixture with newly smoldering charcoals and allow airflow & time to do the rest (See Fig. 4).

Fig. 4 – Smoldering charcoal “bath” for the fixtured NiTi.


Root3 Labs allowed the part shown in the images and video to sit within the smoldering coals for the recommended dwell time of 30-45 minutes, occasionally shifting the coals slightly to maintain smoldering and randomize the heating. At the conclusion of the 30-45 minutes, the part was removed from the charcoals using tongs (please remember the fixture will be HOTTER than a Maryland summer, with an emphasis on the ER!) and immediately quenched in a water bath to “lock-in” the new crystalline structure on a microscale – and the desired shape on a macroscale (Fig. 5 and video).

Fig. 5 – Heated NiTi & fixture prior to quenching.


To sanity check your new cooking skills, remove the quenched part from the water once it has fully cooled to the touch. Give it a few bends. Using the aforementioned tongs – or your new titanium bionic hand compliments of SHOCK Trauma if you skipped the last few sentences – bring the part near the charcoals once again, but do not put it in direct contact this time. If successful, you will witness your creation transform back to the heat-set shape!

Now is an appropriate time to raise your arms and eyes to the sky and unleash a mad scientist’s, “MWUAHAHA!”. No judgment. Recall that this transformation occurs at the lower activation temperature, in our case 176F (80C), hence the reason you only need to bring the part close to the charcoals. Even better, this process can theoretically be repeated indefinitely, although some degradation of the material over many thermal cycles may be possible in the real world.

Some caveats to this process: (1) It is probably definitely not ideal for repeatable, highly controlled, and reproducible parts. For that, just invest in the oven designed for this exact thing and let some other people use the grill for crying out loud! (2) Before all else, consider the part and fixture geometry. For instance, our stock Nitinol sheet was only 0.10in (0.25mm) thick and was uniformly wide (~0.25in [6.35mm]) throughout our part’s profile. On top of that, the fixture that was, well… on top of that… was bulky steel to provide a conductive, but high thermal mass “distributor” between the part and the charcoals to help with heating uniformity. (3) Lastly, the heat-setting temperature range of our Nitinol material had 360F of forgiveness to work with, so obviously the wider the heat-setting range the better for our purposes.

Happy and safe cooking!


Background Section

  1. (2020, October 15). Shape-memory alloy.


  1. Rein, Smoldering Combustion, Chapter 19 in: SFPE Handbook of Fire Protection Engineering, 5th Edition, pp 581-603, Springer, 2016.