" The test is conducted in such a way that neither it will change the failure mechanisms nor introduce new one, but only accelerate existing ones."

if you ensure this point, I think your calculation is acceptable. However, the real situation is that as you increase the cycle frequency, it is very difficult to keep this point; e.g., water hammer is a key effect to influence the mechanical valve behavior, and it should change a lot in the accelerating test. There are quite a lot literature on this topic I think, you may check it and regulate your exp. system carefully. 

Since no way to keep it unchanged, you should focus on some concrete mechanism I think.

Thanks Feng. So the point is valid if the rise is within the safe limits. That means we need to ensure that the rise will not result in any additional failure mechanism.

If it is above the limits and causing new failure, we need to accommodate that physics also to calculate the life. Am I right...

"which means the 10 year TTF at accelerated cycle test became 150 years at real usage conditions (acceleration factor=15)"

This is true, but usually you want to simulate 10 years of operation of valve in normal conditions thus you build the test stand to shorten the time needed for tests. You need to verify if you need to test the valve for 10, 22, 25, 30 etc years or you need to simulate the valve to operate in normal conditions for such a period of time in Time Accelerated Tests. You will find tests that must be performed on valve protheses in standards. You may refer to ISO 5840:2009 Cardiovascular implants. Cardiac valve prostheses in example.

Thanks Damian. Those numbers are just arbitrary values. I have not shared the real values. Tests were conducted based on iso 5840 for less than 10 years.

I was verifying whether the approach i adopted in calculation is correct.

The situation here seems to me a bit more complicated. I must admit that I have no knowledge in heart action and anatomy. But operation of heart valve depends on the mechanical state of the ("artificial") valve itself and on the state of  the ("natural")  living and acting heart  which  may change with time (aging  (!) and may interact with the state of the valve. This interaction may influence, possibly, the valve life duration.

Heart experts could possibly provide a better insight into the situation.

Ilya Gertsbakh  

 basically it is an issue of mechanical similarity.

failure mechanism, the term is too gross, you may care about structural failure, or surface cavitation (for mechanical one), or calcification (tissue valve). so in accelerating exp. impacts on these issues change a lot. since total similarity is hard to be reached, you should focus on particular mechanism and pick out the main infleunce effect to analyze. 

simple exp. may just follow the iso regulatipn on this subject and with a same type commercial product as control. the linear calculation you surposed will be questioned by any reviewer.

Please contact Sree Chitra Tirunal Institute of Medical Sciences & Technology at Trivandrum. They are an institution of national importance under DST, and they carried out these experiments in 1980-83  using a pulse-duplicator and life-cycle testing for the development of the Chitra Heart Valves, later marketed by TTK Pharma.