TI-6-4 heat treatment to minimize grain size 1

[RSEngineer]

I am looking for the optimum heat treatment to make a more homogenous part for an investment cast Ti-6-4 machining blank. My ultimate goal is to make it more machineable for a fine surface finish (<10RMS).

What I have so far is that I think the cycle should be a solution treatment above the beta transus (~1000°C), somewhere like 1050°C, water quench, then age at a lower temp to stress relive, perhaps 650°C for 1-2 hours.

As someone completely new to this field, I would appreciate any advice you can offer.

 

[coolkid]

There is a comprehensive report from the National Physical Laboratory QU 53 May79 by D R Salmon that covers various grades of Titanium - 60 odd pages on Ti6Al4v alone!

&quot;The alloy may be given a variety of heat treatments to produce mechanical properties matching different requirements&quot;
1 Stress relieving - The recommended treatments for stress relieving can vary from 50 hrs. at 870K (597C)to 5 hrs at 925K Partial stress relief (50% reduction in residual stresses)can be achieved by a treatment at 755K (482C)for 1/2 hr.
2 Annealing is generally carried out in the temperature range 980K - 1090K for between 1/2 and 4hrs. followed by either an air or furnace cool depending on the properties required.
3 Solution treating. The heat treatment varies according to the type of properties required. Maximum ductility is obtained by a heat treatment at about 1120K and maximum strength by a treatment in the range 1200-1230K To maximise the response to ageing treatments the material should be rapidly cooled from the solution temperature by an efficient water quench.
4 Ageing treatments are normally carried out in the temperature range 755 - 870K for periods ranging from 1 - 10 hrs. The fracture toughness can be improved at the expense of a lower tensile strength by raising the ageing temperature eg. a four hour treatment at 895K followed by air cooling.

 

[TVP]

Ti6-4 investment castings for aerospace applications are most frequently used in the BSTOA condition-- Beta Solution Treated and Over Aged. This treatment is typically 1 hour at 1024 C, rapid cooling (at least 85 C/min) to a temperature below 250 C, then overaging for 1-2 hours at 843 C. This results in an optimum combination of properties-- fatigue, fracture, etc.

For your application, I would not age at such a high temperature. Aging after water quenching from the beta solution temperature should probably be in the range of 510-550 C. Also, you will want to minimize the time above the beta transus. Perhaps try 30 minutes instead of 60 minutes, although you will need to be wary of incomplete solutioning.

 

[RSEngineer]

Thanks for your help. I am now looking to find the NPL report coolkid referenced. Also, the lower 1024 C temp sounds like a more typical number as I dig into this further, plus the quench to below 250 is a noteworthy addition mentioned by TVP. The notes I have indicate an even steeper quench of 150°C/min. What concerns me is the lower range on the aging, though. I am afraid of residual stresses in the final part. The part will be machined to a wall thickness of about .030 in. When this project was started, the team before me was using powder metallurgy and had been experimenting with times and temperatures for the aging to eliminate an early problem they had with &quot;oil canning&quot;. They were settling on a temperature between 620 and 750 C before we switched to investment castings. Of course, they had gained stresses after they had sintered the powder metal at 1255 C, so they may have had more stresses than I do to begin with.

One other question, would heat treatment to a temp below the beta transus get what I am after? I have seen referenced solution treating and aging to 955 C for 10 minutes, water quenched, then aged for 4 hrs between 540 and 675 C, followed by air cooling to 25 C. Would this get me the better machining part that I am after, or am I better off going above the beta transus for what I am after?

 

[TVP]

I almost suggested annealing in the alpha + beta range, but thought that you may have already investigated that and you really wanted the beta treatment. Before recommending an alpha + beta treatment though, I would want some more detail on your appplication. Environment, loads, etc. Also, details on the casting-- section size (min and max wall thickness), sand type, cooling rate, etc. Beta solution treating is used because investment castings solidfy slowly, which can result in variable grain size, chemical inhomogeneity, etc.

 

[RSEngineer]

A few areas of it are nearly 1/2 in (125mm) thick, but the thickness through most of it is about .170 in (43mm). The application regularly cycles the part in the 150 - 250°F (65-120°C) range and doesn't really have significant pressures or aggressive materials on it, but is cleaned quite regularly by people that are not necessarily the most careful wielding pneumatic drill powered wire wheels and scouring pads.

My investment caster does not like to tell much about the makeup of the sand, something of proprietary data I suppose. Cooling rate is nothing too fast, as evidenced by the large grains we currently experience, but I don't know what the rate is. They just cool it within the skull furnace they make the casting in.

The machining difficulties we have had are what has brought on this project - essentially, the inhomogeneity and inconsistency between batches. I want a more consistent, repeatable, machinable part with a polished surface finish. That is why I am looking at quenching rapidly from somewhere near beta transus to make small grains and a more homogenous grain structure for the machinist.

 

[rnd2]

If you want super rapid cooling, could you quench into a liquid metal or combination of, say aluminium, tin or lead?

 

[TVP]

RSEngineer,

Based on your latest information, I would recommend homogenizing above the beta transus. Keep the temperature and time as low as possible, water quench with good agitation, and then age somewhere between 510 and 675 C. 4 hours is quite a long time to age non-aerospace parts. Good luck.

 

[NickE]

In response to rnd2: I thought that the quickest possible quench rates were obtained using Agitated brine? Unless liquid metal conduction rates are so very fast that the part cools to the Mp of the quench medium faster than water can flash to steam.

Another related question: Is the Cp of liquid metals generally greater than the enthalpy of vaporization of water? IE: if the amount of heat absorbed by vaporizing water is lower than the Cp of (lead/tin etc) then I could easily see that liquid metal quenches would be faster.

I thought that liquid metal (salt pot too) quenches were for &quot;quench to temp&quot; precision and to prevent quench cracking by avoiding the last bits of dimensional change.



Nick
I love materials science!

 

[rnd2]

Nick
As the hot part is quenched, heat is transferred to the water next to it, in turn evolving into steam (a gas). A gaseous nexus is created next to the part and thus works as an insulator. Moving the part or the fluid rapidly while water quenching reduces the nexus and speeds cooling but runs the risk of creating diffential surface cooling. To my knowlege nothing &quot;cools&quot; faster than immersion into liquid metal. The trick is to provide liquid at the desired temperature.

 

[crackclosure]

try this link &quot;

http://www.eng-tips.com/gviewthread.cfm/lev2/15/lev3/55/pid/367/qid/62749&quot;

it may helpful to you.....

 

[unclesyd]

Here is a paper you can get for US $15.00 at

http://www.doc.tms.org

It's in the October issue




Microstructure Evolution during Alpha-Beta Heat Treatment of Ti-6Al-4V [pp. 2377-2386]
S.L. Semiatin, S.L. Knisley, P.N. Fagin, F. Zhang, and D.R. Barker


--------------------------------------------------------------------------------
A framework for the prediction and control of microstructure evolution during heat treatment of wrought alpha/beta titanium alloys in the two-phase field was established via carefully controlled induction heating trials on Ti-6Al-4V and accompanying mathematical modeling based on diffusioncontrolled growth. Induction heat treatment consisted of heating to and soaking at a peak temperature Tp = 955 °C, controlled cooling at a fixed rate of 11 °C/min, 42 °C/min, or 194 °C/min to a variety of temperatures, and final water quenching. Post-heat-treatment metallography and quantitative image analysis were used to determine the volume fraction of primary (globular) alpha and the nucleation sites/growth behavior of the secondary (platelet) alpha formed during cooling. The growth of the primary alpha during cooling was modeled using an exact solution of the diffusion equation which incorporated diffusion coefficients with a thermodynamic correction for the specific composition of the program material and which took into account the large supersaturations that developed during the heat-treatment process. Agreement between measurements and model predictions was excellent. The model was also used to establish a criterion for describing the initiation and growth of secondary alpha as a function of supersaturation, diffusivity, and cooling rate. The efficacy of the modeling approach was validated by additional heat treatment trials using a peak temperature of 982 °C.