Hveem stablity too low but Marshall stability is good, how is the mix?

[hobbie]

We have a job that requires Hveem stability of 37 minimum. The mix used 25% RAP and the production was on the high side of AC content (5.6% vs the 5.0% as target), the air voids dropped to around 1% and Hveem stability was only 12. The lab mix design data with around 5.6% AC shows around 30 stability, and has air voids about 2.5%. However, we verify the AC content was indeed only 5.6%. The air voids might be closed by extra breakdown from plant production.

Since we never saw a Hveem stability that low with our material, I suspect it might be due to the sensitivity of Hveem test method to the low air voids. Just to verify this thought, we also tested the leftover material from the QC with Marshall compaction. The result verified the low air voids, however the Marshall stability was very good (4500-5000 lbs) with high flow (18-20 units). If the material fails at the half the curve, it will be 2500 lbs of stablity and 10 of flow, and this result will meet most FAA specification. I personally think the materials with 5000lb stabilty and 20 flow will out perform the one has less stabililty and flow. What do you guys think?

As part of investigation, I was hoping that we can run some tests such Hveem stability and Marshall stability directly on the 4" field cores, then I just found out the diameter of the 4" field cores are actually around 3.75 inch. Any suggest on that? Can we still test these cores?

The total materials paved is about 200 tons of dense graded HMA. It will be covered by another 2 inches HMA on top it later before put in service. The higher potential for rutting may not be much of the concern since I remember I read somewhere the rutting happens mainly on the surface layer. The application is an entrance loop for a parking structure.

The bottom line is that we have concern on whether we can keep it in place with additional warranty or we should just remove it to prevent potential performance issue in the future. We may need some good support to convince the owner for us to go with the warranty option as well.

 

[Ron]

Since you ran the Marshall stability from left over material, that would suggest that you re-heated the sample. Marshall stability goes up tremendously on a re-heated sample. I would not consider those values to be valid. A stability of 4500 with a flow of 18-20 is an indication of a problem with the test.

The cores can be tested, but just not for stability in the Marshall apparatus. You can do a splitting tensile test using the Marshall apparatus and then correlate that. There are published correlations. I know a correlation study was done at Clemson University in the early 1980's.

Get more cores if you want to test the in-situ Marshall value. Get a core barrel that has an ID closer to 4.0 inches and use flexible shims if necessary. You'll have to adjust your values if you use shims.

Have you done an extraction and gradation to check the mix?

 

[hobbie]

Ron, thank you very much for the response. We did reheat the sample for Marshall test, however, the reheat was tightly controlled just to get materials loose and ready for compaction. The stability value from reheat sample is similar or just slightly higher than the lab design value. We do have a similar mix (virtually the same gradation and same binder) design by Marshall method, and the Marshall stability of the mix design was between 4000-4500, and flow around 15 with at 5.5%AC. The reheating might increase the Marshall stability, but not a lot.

If we consider the Marshall test result, what do you think the mix? Is it good enough to stay or still have some performance concern with the high flow and low air voids?

We just get a couple of slabs from the field, my plan is to test the AC, gradation, and compact a few more pills by both Hveem and Marshall. Any suggestion on that?

BTW, when you refer to the splitting tensile test, do you actually mean indirect tensile test, just as the TSR testing? Any hint to find that literature about the Clemson University will be highly appreciated.

 

[Ron]

If you consider only the Marshall results, the mix is quite good. We generally see Marshall stabilities in the 2000 to 3000 range, with flow in the 10 to 12 range.

I think you need to validate the difference between the Hveem and Marshall values. Something is quirky there. The testing program you contemplate will likely show the anomaly.

 

[hobbie]

I just did some Indirect tensile strength on these 3.75" diameter cores since I don't have any good use of these cores. The strengh at 77F degress was very impressive, in the range of 280-320psi. Our typical dry strengh as part of TSR test usually below 150psi.

We was planning to send these 6" cores out for APA rutting test, but the diameter was about 5.6" instead, not sure if the APA can accomodate that?

 

[Drumchaser]

I apologise that I evidently have nothing constructive to add, but would like to chime in if possible.

The 25% Rap mixture - Was the Rap fractionated at the plant site?

As stiff (tough) as our Rap mixtures have been, I would not be worried about rutting, based on the empirical data you have noted above.

 

[Ron]

I agree with Drumchaser about the rutting. The mix seems to stiff to induce rutting in the asphalt layer. Rutting is usually a function of subgrade deflection, not asphalt, unless you have an unstable mix.