Mantenimiento Predictivo - New Challenges in the European Area

New Challenges in the European Area:
Young Scientist’s 1st International Baku Forum
Baku (Azerbaijan) - May 20th-25th 2013
MANUEL SOBRINO
University of Valladolid (SPAIN)
The future of liquid fuels:
viscosities at high pressures
OUTLINE
1.- Introduction and objectives
2.- Experimental setup and procedure description
3.- Viscosity model
4.- Calibration procedure
5.- Verification of the calibration
6.- Evaluation of the uncertainty
7.- Conclusions
1.- Introduction and objectives

1.1.- What are we looking for?

We are trying to “find a solution” for the problem of viscosity at high pressure…

Which is that problem?

Limited experimental data

High uncertainties (oldest data with large deviations)

Lack of models and measurement techniques
1.- Introduction and objectives

1.2.- What does “find a solution” mean?

Develop an accurate technique which allow us to get viscosity data at high pressures

Anaylize the literature in order to see what is happening and how to improve…

Find a model which completely describes our viscometer behaviour

Determine the real uncertainty of the apparatus taking into account all components
1.- Introduction and objectives

1.3.- Viscosity measurement techniques

Four main techniques:

Oscillating body viscometers

Capillary viscometers

Vibrating viscometers

Falling body viscometers
Vibrating wire viscometer
FALLING CYLINDER VISCOMETER
2.- Experimental setup and procedure description

2.1.- Working principle

The longer the duration of the body fall is, the higher the viscosity is (up to 140 MPa)
Falling body
High pressure tubes
Detection coils
2.- Experimental setup and procedure description

2.2.- Experimental setup
2.- Experimental setup and procedure description

2.2.- Experimental setup

Measuring cell (TOP Industry – designed by University of Pau, France)
2.- Experimental setup and procedure description

2.2.- Experimental setup

Thermostatic Bath (Julabo F81-ME)
2.- Experimental setup and procedure description

2.2.- Experimental setup

Pressure generator (HiP 50-5.75-30) and digital gauge (Druck DPI 104)
2.- Experimental setup and procedure description

2.2.- Experimental setup

Vacuum pump (Leybold TRIVAC D8B) with a cold trap
2.- Experimental setup and procedure description

2.2.- Experimental setup

Electronic devices
Vacuum display
Data Acquisition Unit
Agilent U2352A
Waveform Generator
Agilent 33220A
Temperature measurement
Agilent 34970A
2.- Experimental setup and procedure description

2.3.- Procedure description

Computer program using Agilent VEE Pro software: body detection (d/D>0.93)
3.- Viscosity model

3.1.- Models used by other authors

Expression widely used:
being
(tendencies…)
3.- Viscosity model

3.2.- New model

Previous practical work: improvment of the experimental design made in Pau

Relationship which includes all characteristic variables:

It can also be expressed as follows:
4.- Calibration procedure

4.1.- First step: minimizing punctual errors

Toluene: 273.15 K to 353.15 K and 0.1 MPa to 140 MPa

Dodecane: 273.15 K to 313.15 K and 0.1 MPa to 140 MPa
4.- Calibration procedure

4.2.- Second step: proposed model
5.- Verification of the calibration

5.1.- Verification with heptane
Heptane at 293.15 K
Deviations within ± 3.5 %
Heptane at 313.15 K
6.- Evaluation of the uncertainty

6.1.- Monte Carlo simulations


Two main ways to evaluate the uncertainty:

Rigorous procedure of EA-4/02 guide (variance propagation law)

Monte Carlo (for more complex equations, random numbers)
Results of three simulations:

0.5518 ± 0.030 mPa·s (± 5.44 %)

1.124 ± 0.032 mPa·s (± 2.85 %)

2.299 ± 0.045 mPa·s (± 1.96 %)

Uncertainty less than ± 3 %

Normal distribution (95.45 %)
7.- Conclusions and future work

7.1.- Conclusions

A bibliographic study has been done to check the state of art and the problems

A model to describe the complete behaviour of our viscometer has been proposed

A calibration of the equipment has been done with toluene and dodecane

This calibration has been verified with heptane

Uncertainty calculations have been carried out using Monte Carlo method (± 3 %)
FINAL CONCLUSION: THERE IS STILL A LOT OF WORK TO BE DONE IN THIS FIELD!
Acknowledgements

Organizers: Eurodoc & Ministry of Youth and Sports of Azerbaijan Republic

Dr. José Juan Segovia

TERMOCAL Group

Spanish Government - Science and Innovation Ministry (ENE2009-14644-C02-01)
AND OF COURSE…
THANK YOU FOR YOUR ATTENTION!