Lime for Bedding

Q I have used a specified mix of 3 parts sharp sand (with no less than 40% quartz chippings) to one part Hydraulic lime on a medieval building in Oxford and have come up against a problem with the floor.
After 2 years of the floor being laid, the stone floor slabs have become loose and unbonded and the mixture beneath is soft and crumbling. The floor has been subject to under-floor heating and I am assuming this would be the probable cause for the breakdown in the bedding. The floor was laid (40 – 50mm thick bedding) and the specification was to prevent heat being introduced for 7 days, and then to gradually introduce heat after the 7-day period but not to raise the heat above 19 degrees for the next month to achieve initial set. I believe the floor was subjected to heat sooner than the 7-day mark at temperatures higher than 19 degrees and this allowed the lime mortar to dry out before carbonation took place, affecting the integrity of the mixes strength. Since then I have carried out some remedial work and was able to leave the heat off the floor for approximately one month. Added to this I included the potzolan additive Metastar to aide in the curing time because of the buildings need for public use the month after the floor was laid.The remedial work has shown no signs of movement and appears to have worked.
In a professional capacity, is it fair to say the heating is the cause for the breakdown in the strength?
The client is thinking of investigating the mixture but because I have changed the mix by adding metastar, do you think conclusions will be met by testing the mix and what should be expected from the investigation outcome?
Can they tell me if the mix was correctly mixed?
Can they tell me if the Lime was suitable in the mix?
Can they tell me if the lime has carbonated or not by testing it?
I would be grateful if somebody could explain what can be done to conclude the mystery of why the mix has failed in this manner.

A It is not really possible to give definitive answers to your questions without further investigations, but you are correct in saying that the heat would cause excessively quick drying of lime and thus prevent carbonation. However, there are one or two other points you need to take into account:
Laying stone slabs over a lime bed will also prevent carbonation, as the stone will only allow carbon dioxide to penetrate through the joints: almost certainly not enough to permit carbonation of any more than a tiny fraction of the bedding mortar. Assume therefore that an effective set based on carbonation cannot in practice happen in this instance.
In this situation it is therefore correct to use a hydraulic lime, which will set in the absence of CO2. You do not say which type of hydraulic lime you use, but a 1:3 mix of hydraulic lime is slightly weak in itself. Your addition of pozzolan will have increased the effective strength of the hydraulic set by combining with the free lime which in this instance is probably not carbonating anyway. Even with hydraulic lime, the longer you can allow for curing the better, and the application of any artificial heat is usually detrimental.
What can be discovered through testing depends on the tests applied, some of which can be very sophisticated/expensive, especially when analysing hydraulicity. Interpretation of results will still be necessary.

A Other possible causes of the problem encountered are:
Inadequate wetting of the surface onto which the mortar was laid and inadequate wetting of the slabs prior to laying. It would normally be desirable to thoroughly wet the whole base a few hours prior to laying with a further wetting of smaller areas and the slabs immediately prior to laying. Without this the suction of the dry base and slabs would cause the mortar to dry out far too quickly leading to its failure. The application of heat would accentuate this. Even with the wetting as above it may still be necessary to further wet the slabs after laying.
How was the mortar mixed? If a drum mixer was used it is recommended that 20 minutes should be allowed to ensure a thorough mix; preferably mixing the sand and lime dry before adding any water.
As to the application of heat, I would expect any within the 30 days after laying to be detrimental, and it ideally should be left longer.
An excellent booklet is Technical Advice Note 1, ‘Preparation and Use of Lime Mortars’ written by the Scottish Lime Centre Trust for Historic Scotland (available through the Bookstall).