pietkoning
Mechanical
- Sep 11, 2014
- 7
Dear community members,
After having searched in several books and on the internet for a good reference, I must conclude that a good description to assess measured vibrations is difficult to find.
Suppose the goal is to perform a pipe vibration measurement and subsequent assessment. The following steps apply to my opinion:
1) measure the velocity of the pipe, resulting in a graph with velocity vs. time (time on horizontal axis)
2) perform a Fourier transform on the measurement results, resulting in a graph with velocity vs. frequency (freq. on horizontal axis)
3) assess the result from (2) using a standard as VDI 3842 or equivalent.
The question is how to perform the 3rd step. As an example I have added the result of a Fourier transform as per step (2) and an assessment graph like in VDI 3842 and comparable codes.
Figure (A): Result from step (2): Link
Figure (B): Graph like in VDI 3842 as used in step (3): Link
You may lay the VDI 3842 graph lines in figure (B) on top of figure (A), but I wonder if this is correct. My doubt is for the following reason:
case (I): suppose there is one measured frequency f1 = 10 Hz with a corresponding velocity of 0.1 mm/s. Plotting this INDIVIDUAL point in the assessment graphs (figure (B) ) shows that you are below the pink line, indicating you are somewhere below 'strong vibrations'.
case (II): suppose there are two measured frequencies f1 = 10 Hz and f2 = 20 Hz, both with a corresponding velocity of 0.1 mm/s. If you plot both points INDIVIDUALLY in figure (B), then both points are below the pink line. However, the vibration in case (II) is clearly worse than in case (I). Can you just plot both points individually or not?
case (III): suppose the most general situation of a vibration measurement with the result of a Fourier transform as in figure (B). Like in case (II), you might wonder if you can just lay the assessment lines in figure (B) over the result of the Fourier transform as in figure (A). Case (III) is just a generalization of case (II), since you have different vibrations running through each other, each with another frequency and a corresponding velocity.
Maybe the assessment graphs are generally so conservative that you can just lay the assessment graph lines over the Fourier transform results (i.e. lay Figure (A) over Figure (B)), but it is never stated if that is the case.
Anyone who has experience with performing such an assessment or who has otherwise an opinion about this interesting topic?
After having searched in several books and on the internet for a good reference, I must conclude that a good description to assess measured vibrations is difficult to find.
Suppose the goal is to perform a pipe vibration measurement and subsequent assessment. The following steps apply to my opinion:
1) measure the velocity of the pipe, resulting in a graph with velocity vs. time (time on horizontal axis)
2) perform a Fourier transform on the measurement results, resulting in a graph with velocity vs. frequency (freq. on horizontal axis)
3) assess the result from (2) using a standard as VDI 3842 or equivalent.
The question is how to perform the 3rd step. As an example I have added the result of a Fourier transform as per step (2) and an assessment graph like in VDI 3842 and comparable codes.
Figure (A): Result from step (2): Link
Figure (B): Graph like in VDI 3842 as used in step (3): Link
You may lay the VDI 3842 graph lines in figure (B) on top of figure (A), but I wonder if this is correct. My doubt is for the following reason:
case (I): suppose there is one measured frequency f1 = 10 Hz with a corresponding velocity of 0.1 mm/s. Plotting this INDIVIDUAL point in the assessment graphs (figure (B) ) shows that you are below the pink line, indicating you are somewhere below 'strong vibrations'.
case (II): suppose there are two measured frequencies f1 = 10 Hz and f2 = 20 Hz, both with a corresponding velocity of 0.1 mm/s. If you plot both points INDIVIDUALLY in figure (B), then both points are below the pink line. However, the vibration in case (II) is clearly worse than in case (I). Can you just plot both points individually or not?
case (III): suppose the most general situation of a vibration measurement with the result of a Fourier transform as in figure (B). Like in case (II), you might wonder if you can just lay the assessment lines in figure (B) over the result of the Fourier transform as in figure (A). Case (III) is just a generalization of case (II), since you have different vibrations running through each other, each with another frequency and a corresponding velocity.
Maybe the assessment graphs are generally so conservative that you can just lay the assessment graph lines over the Fourier transform results (i.e. lay Figure (A) over Figure (B)), but it is never stated if that is the case.
Anyone who has experience with performing such an assessment or who has otherwise an opinion about this interesting topic?
