excel chart equation

[coopdog]

I have a question regarding the excel trend line and he equation it provides.

I have a temperature vs viscosity graph and I have added a eponential trend loine and displayed the equation. I have used that equation to calculate actual readings vs where they should fall on my graph. The equation works very well. What my question is does anyone know how to do the following.

I have a viscosity spec at a specific temperature, if I use my equation it will tell me what the spec should be at my higher temperature. If I take a reading at say 10 deg higher than what my spec is, and the reading is say 5 points higher than what my trend line equation says it should be, how can I calculate how over spec that 5 points is over my lab spec at the lower temperature. It is not a linear line so I do not think 5 points over at one temperature will be 5 points over at another.

Any Ideas, this may be more of a math question than an excel but I figure there are lots of smart people around this forum.

Thanks

Mike

 

[Jon Peltier]

If you have a formula for viscosity vs. temperature, and you can relate it to the data point you have at a too-high test temperature, you could use the relationship to determine what the measured viscosity would be at the spec temperature.

You didn't post the formula nor say how well it fits, so nobody can tell you more than that. If you post the trendline formula and the data that you fitted to attain that formula, I or someone else might be able to make sense of it.

 

[coopdog]

y = 185.28e-0.0178x is the equation, and it fits very well, also if I calculate the values out with this formula it represents fairly close to what I am seeing online.

thanks for the reply.

Mike

 

[Jon Peltier]

If that's the formula for the spec behavior, then at the indicated temperature it should produce the target value of viscosity. Enter the off-spec temperature, and it will tell you the corresponding target viscosity, which you can compare to the value you measured.

If the formula describes your measured viscosity, then plug in the actual spec temperature to estimate your viscosity at the spec temperature.

Would your formula fit better if you used absolute temperature? Viscosity and other rate phenomena that fit an exponential temperature relationship are fitted using kelvin, rather than celsius.

 

[coopdog]

Yes that works very well. The issue I have is that at the spec visc there is a tolerance of +-5, is there a way using this formula to estimate if that +-5 is the same amount of change while moving to a higher off spec temperature. If at a higher temperature it is 5 seconds high on visc based on this formula's estimate if I were to cool it down the the correct temperature would it be 5 seconds high, or would it be 2 seconds high or 10.

Thanks again for the help.

Mike

 

[Jon Peltier]

Your formula is
y = 185.28e-0.0178x

You need more information to answer your question. I cannot help you, because I've never studied viscosity of materials.

Do you have a formula for the +5 tolerance and the -5 tolerance? In other words, is the difference taken up by changes to the pre-exponential term or by the exponent? (Without knowing exactly, I would think the pre-exponential would have more variability.) Do you have any theoretical work that points to an answer? These are the kinds of question you need to ask.

 

[coopdog]

thanks anyway I will post if I figure it out.