HDT test - deflection v. time curves 1

Hello Entanglement,

HDT and Vicat machines now come with the data acquisition that you are looking for, contact Tinius Olsen Testing Machine Co. ( for more information and a quotation.

All the tests that I have seen, (less than 50), have shown a non-linear curve. More and expotential function than a straight line. I would imagine you cna get more information from some of the on-line sources and direct from the producer, ask your supplier rep and he might be able to get the information for you.

Good luck

Greg Thelen

Entanglement:

Heat deflection versus time curves may not be particularly useful. The test apparatus should have a linear rise in temperature, if it’s calibrated correctly. However, the temperature of the specimen is anything but linear.

Because polymers are natural insulators, they are poor conductors of heat. Therefore, the outer surface heats more quickly than the inner core. The specimen effectively becomes a “composite” specimen with a soft outer periphery and a hard, reinforcing inner core. This “composite” structure of the specimen changes in a very nonlinear fashion depending on the shape, primarily the thickness, of the specimen and the thermal conductivity of the material.

To compound the error, most specimens have residual stress which can be relieved at a certain temperature. This stress relief can exhibit a sharp drop off in deflection, or if the specimen is inverted, the deflection may begin to “decrease”, i.e., the specimen may deform upward against the load as it releases the frozen-in stress. This will raise serious havoc with your data.

So be careful using HDT data for this purpose. You may want to perform long-term elevated-temperature creep studies to develop the information you’re seeking.




Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

Thank you for your valuable posts.

Heat deflection versus time curves may not be particularly useful.

Click to expand...

So be careful using HDT data for this purpose. You may want to perform long-term elevated-temperature creep studies to develop the information you’re seeking.

Click to expand...

I definitely agree about the lack of use of the HDT test data, since a lot of errors affect it (I never thought of the facts you point out though - thanks for that).

Unfortunately HDT is very often (if not always) the only information about the thermal behaviour of a specific polymer grade that can be obtained by most polymer producers... Actually I wonder why even BIG companies keep on sticking on it, instead of switching to some more significant test i.e. creep tests at high temperature as you suggest...

The main reason why I look at those data is to compare different materials to choose the best for my application. The problem is I don't have the testing machines required to obtain such data by myself - and I think the request to polymers producers for a high temperature creep curve could just generate panic ! how about requesting a master curve and shift factors? Chances are that the people I am asking to don't even know what I am talking about... and even if they knew it, would they conduct those tests just for me ? Looks like the rest of the world is OK with HDT...

I was wondering how to improve the number of informations that can be obtained by HDT tests then... if the HDT value (single value) is almost no use, "Heat deflection versus time curves may not be particularly useful" but it's still an improvement, although small!

At least if the testing machines used for HDT test actually record the whole curve I don't think it would be too difficult to provide that - while a creep test would require different testing machine and different experience...

Best regards,
Sergio

Sergio:

There is nothing wrong with using the HDT to differentiate between different candidate materials. You just have to remember that HDT is just one test, which has its strengths and limitations. Understand the test and how the data is developed and extract the useful information that can be gleaned from it. Do not, however, make your judgement on that number alone --- and definitely not on the value that is published. Published data sheet properties should be used as indicators to help you weed through the multitude of possible candidates and eliminate those materials that just won’t work and limit the ones that require further investigation.

The fact that you don’t have “sophisticated test machines” doesn’t mean you can’t do your own testing. If elevated-temperature creep is the data that you need, you can simply take tensile bars, mark two lines on each, hang them in an oven at an appropriate temperature, hang a weight on them, and periodically measure the dimension between the lines. Plot the data versus the time in the oven and you will quickly begin to generate long-term elevated-temperature creep data that is specific to your needs. Not only that, but in short order you will begin to get a “feel” for the different behavior of a number of materials that you can’t get from reading data sheets. With time your knowledge of material behavior will be invaluable to both you and your company.

The thing most people don’t understand is that long-term testing doesn’t take as long as you think. By using time on the log scale, most of your curve is developed in a few weeks. All you need is a few data points that continue for the long term --- and those you only have to look at periodically. So go ahead. Take the step. Start some long-tem testing. You’ll be surprised at how much you’ll learn.




Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

RichGeoffroy said:

There is nothing wrong with using the HDT to differentiate between different candidate materials.

Click to expand...

My main concern about HDT test is that since its specifications are just conventional and arbitraty it fails to diffentiate between materials: depending on the conventional maximum allowed deflection material A can be better than B or just the opposite. HDT is not a material property as e.g the creep compliance v time curve.
Moreover since a lot of errors can affect the measure - as you pointed out very clearly - I think the comparison is even less reliable.

However I agree with you on the fact that those data should be used only as a first rough approximation - actually also that rough comparison could be wrong though!!!
Just I would be a lot happier if the materials producers gave at least the modulus v temperature curve of their grades!

Best regards,
Sergio

Have you thought of using DSC (differential scanning calorimetry) and look at the onset of the Tg? It would not be subject to the concerns raised by Rich Geoffroy. This is what I usually use as a comparator to HDT. I am working with thermosetting plastics.