Dear Experts 

i face a problem at a raw vertical mill MAAG Gear model WPU (forcing frequencies are attached at the file)

Gearbox type : Epicyclic gearbox 

problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

Attachments

Original Post
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

John from PA posted:
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear.

i have doubts also in the concrete or the soil under the whole structure ,beaneath the concrete.

Hossam 

 

Hossam 2 posted:
John from PA posted:
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear.

i have doubts also in the concrete or the soil under the whole structure ,beaneath the concrete.

Hossam 

 

Give us a number!

John from PA posted:
Hossam 2 posted:
John from PA posted:
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear.

i have doubts also in the concrete or the soil under the whole structure ,beaneath the concrete.

Hossam 

 

Give us a number!

On load vibration fluctuates from 3 to 8 mm./sec thus it stops the mill

While on. No load vibration. 0.5 mm. /sec

Same mills on another plants operates between 1.5 to 2.0 mm/sec on Load condition

By number, I was referring to something like 16.5 Hz, your bevel pinion speed.

None of your plots are labeled as to loaded or unloaded.  Please attach the plot that shows what you are talking about.  To minimize confusion, attach ONLY that plot, none of the other plots.

Hossam 2 posted:
John from PA posted:
Hossam 2 posted:
John from PA posted:
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear.

i have doubts also in the concrete or the soil under the whole structure ,beaneath the concrete.

Hossam 

 

Give us a number!

On load vibration fluctuates from 3 to 8 mm./sec thus it stops the mill

While on. No load vibration. 0.5 mm. /sec

Same mills on another plants operates between 1.5 to 2.0 mm/sec on Load condition

hi hossam

does your soil composition and aggregation is same to another plants?

and how about mass and speed of soil in input mill? are the same to another

John from PA posted:

By number, I was referring to something like 16.5 Hz, your bevel pinion speed.

None of your plots are labeled as to loaded or unloaded.  Please attach the plot that shows what you are talking about.  To minimize confusion, attach ONLY that plot, none of the other plots.

I didn't attach any plots on NO Load condition all of them loaded  condition.

Input speed 998 cpm /16.5 hz 

I didn't mentioned numbers as it was in the attached file. 

Sorry for misunderstanding by me 

 

aziz58 posted:
Hossam 2 posted:
John from PA posted:
Hossam 2 posted:
John from PA posted:
Hossam 2 posted:

Dear Experts 

...problem existence is only on load when the rollers at the grinding table , a unique ferquency shows and its amplitudes fluactuates reaching the alarm level and the shutdown level ,this frequency has a broadband energy  and such frequency dosent related to the forcing frequencies at all at the attached file.

thanks for help 

 

How about giving us a clue as to what that unique frequency might be?

this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear.

i have doubts also in the concrete or the soil under the whole structure ,beaneath the concrete.

Hossam 

 

Give us a number!

On load vibration fluctuates from 3 to 8 mm./sec thus it stops the mill

While on. No load vibration. 0.5 mm. /sec

Same mills on another plants operates between 1.5 to 2.0 mm/sec on Load condition

hi hossam

does your soil composition and aggregation is same to another plants? I thinls soil is different from place to another 

and how about mass and speed of soil in input mill? are the same to another

I have no idea what does it mean.. Mass and speed of soil in input mill? 

 

Hossam 2 posted:
John from PA posted:

By number, I was referring to something like 16.5 Hz, your bevel pinion speed.

None of your plots are labeled as to loaded or unloaded.  Please attach the plot that shows what you are talking about.  To minimize confusion, attach ONLY that plot, none of the other plots.

I didn't attach any plots on NO Load condition all of them loaded  condition.

Input speed 998 cpm /16.5 hz 

I didn't mentioned numbers as it was in the attached file. 

Sorry for misunderstanding by me 

 

What is the frequency (in Hz) that you are concerned about?  Is it the 998 cpm/16.5?  Is it some other frequency?

John from PA posted:
Hossam 2 posted:
John from PA posted:

By number, I was referring to something like 16.5 Hz, your bevel pinion speed.

None of your plots are labeled as to loaded or unloaded.  Please attach the plot that shows what you are talking about.  To minimize confusion, attach ONLY that plot, none of the other plots.

I didn't attach any plots on NO Load condition all of them loaded  condition.

Input speed 998 cpm /16.5 hz 

I didn't mentioned numbers as it was in the attached file. 

Sorry for misunderstanding by me 

 

What is the frequency (in Hz) that you are concerned about?  Is it the 998 cpm/16.5?  Is it some other frequency?

Another frequency it's almost 680 cpm or 11. 3 hz 

Thanks 

I don't see anything of concern at those specific frequencies of 990 cpm and 680 cpm.  Your warning, if I interpret the plots correctly is at 6 mm/sec and these frequencies are well under 3 mm/sec.  What exactly is your concern?

By the way, to look at a machine presentating data like you've shown, I would not have a plot out to 60000 cpm.  This makes your 1X bevel input appear in the far left portion of the plot, less than 2% of the plot scale.

 

John from PA posted:

I don't see anything of concern at those specific frequencies of 990 cpm and 680 cpm.  Your warning, if I interpret the plots correctly is at 6 mm/sec and these frequencies are well under 3 mm/sec.  What exactly is your concern?

By the way, to look at a machine presentating data like you've shown, I would not have a plot out to 60000 cpm.  This makes your 1X bevel input appear in the far left portion of the plot, less than 2% of the plot scale.

Then I can post some other plots to zoom in the band of concern, what should I show too?

Thanks 

what has been the behavior over time of this machine?  Is this a new problem, a problem that perhaps had never been observed in the past, or one that occured after some form of mechanical maintenance?

On scaling, these are just guidelines.  What you look at should be based on the type of machine.

In general for a gearbox look out to about 2X+ of mesh.  Some people recommend 3X+ of mesh.  That would mean for this instance look out to about 600 Hz (36000 cpm) or 850 Hz to about 900 Hz.  This would be your already provided plots out to 60000 cpm.  I do see in one of the plots, something of low magnitude at around 16000 cpm, which might be the bevel mesh frequency.

Once you've looked out far enough to see mesh, 2X mesh, and perhaps 3X mesh, then pull in your sample frequency to look at the rotational components and some harmonics, perhaps out to about 5X.  Your highest rotational is the bevel pinion at 990 cpm, so a plot out to about 5000 cpm would be desirable.

But your have to keep in mind the machine design.  The sun pinion rotational is at 4.26 Hz or about 256 cpm, so it might be desirable to have a plot out to about 1000 cpm, roughly 4X of the sun pinion.

On scaling amplitute, I also suggest you scale everything the same.  Currently your plots have amplitude scale ranging from 1.2 to 3.5 mm/sec.  If everything was scaled to 3.5 mm/sec, it would be readily apparent which physical position had the highest level of concern.

Hopefully you can see that although one plot scale may get you the overall picture, you may have to zoom in on certain areas for diagnostic purposes.

 

 

 

 

 

 

 

John from PA posted:

what has been the behavior over time of this machine?  Is this a new problem, a problem that perhaps had never been observed in the past, or one that occured after some form of mechanical maintenance?

not new problem

On scaling, these are just guidelines.  What you look at should be based on the type of machine.

In general for a gearbox look out to about 2X+ of mesh.  Some people recommend 3X+ of o about 900 Hz.  This would be your already provided plots out to 60000 cpm.  I do see in one of the plots, something of low magnitude at around 16000 cpm, which might be the bevel mesh frequency.

 

Once you've looked out far enough to see mesh, 2X mesh, and perhaps 3X mesh, then pull in your sample frequency to look at the rotational components and some harmonics, perhaps out to about 5X.  Your highest rotational is the bevel pinion at 990 cpm, so a plot out to about 5000 cpm would be desirable.

But your have to keep in mind the machine design.  The sun pinion rotational is at 4.26 Hz or about 256 cpm, so it might be desirable to have a plot out to about 1000 cpm, roughly 4X of the sun pinion.

On scaling amplitute, I also suggest you scale everything the same.  Currently your plots have amplitude scale ranging from 1.2 to 3.5 mm/sec.  If everything was scaled to 3.5 mm/sec, it would be readily apparent which physical position had the highest level of concern.

Hopefully you can see that although one plot scale may get you the overall picture, you may have to zoom in on certain areas for diagnostic purposes.

Ok i Agree with all of that but there is no relationship between the frequency shown in the spectrums with forcing frequencies from the manufacture.?

 

 

 

 

 

 

 

 

Hossam 2 posted:
 

Ok i Agree with all of that but there is no relationship between the frequency shown in the spectrums with forcing frequencies from the manufacture.?

With the frequency scale you've been using (60000 cpm) it remains somewhat a guessing game to put a numerical value on the frequency peaks.  Hence, difficult to establish a probable source.

Untitled

Attachments

Photos (2)

"this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear."

Have you determined the rotational frequency of the mill's balls/rollers and table? I have worked on a coal pulverizer, and saw two frequencies that were very dominant: roll speed and table segment frequency (#segments x table speed and at x #rollers). The mill had a constant (nearly) speed induction motor, but there was no roll vibration when there was no coal in mill, and the roll speed varied depending upon the amount of coal in the mill. I recall the roll speed dropped as more coal was in the mill. I attributed this effect to slip and greater equivalent radius of the rollers. In the coal mill the table rotated, and the roller axis was fixed. You may be measuring vibrations on gearbox, but some mill vibration frequencies will also be present. Perhaps this little story is not relevant to your situation!

Walt

John from PA posted:
Hossam 2 posted:
 

Ok i Agree with all of that but there is no relationship between the frequency shown in the spectrums with forcing frequencies from the manufacture.?

With the frequency scale you've been using (60000 cpm) it remains somewhat a guessing game to put a numerical value on the frequency peaks.  Hence, difficult to establish a probable source.

Untitled

11.38 hz(678 cpm) and its harmonic 22.76 hz (1365 cpm )

thanks 

Walt Strong posted:

"this unique frequency is load dependent frequency ,, as the grinding rollers are in load position this frequency shows and fluctuates ,, by the time the grinding rollers are in the upper position (no load) this frequency disappear."

Have you determined the rotational frequency of the mill's balls/rollers and table? I have worked on a coal pulverizer, and saw two frequencies that were very dominant: roll speed and table segment frequency (#segments x table speed and at x #rollers).

no of rollers 3 and table speed is 26.4 cpm but what is the segments.?

the rollers is free to move i dont think they have a certain speed and its speed depends on the grininding condition.

The mill had a constant (nearly) speed induction motor, but there was no roll vibration when there was no coal in mill, and the roll speed varied depending upon the amount of coal in the mill. I recall the roll speed dropped as more coal was in the mill. I attributed this effect to slip and greater equivalent radius of the rollers. In the coal mill the table rotated, and the roller axis was fixed. You may be measuring vibrations on gearbox, but some mill vibration frequencies will also be present. Perhaps this little story is not relevant to your situation!

the concerning frequency is 670 cpm and the table speed is 26.4 cpm the number of rollers is 3 

Walt

 

What is the roll speed? From geometry the roll speed can be estimated by the ratio of table diameter (at point of roll contact) to roll diameter times table speed. I assume roll diameter is less than table rotation speed, so rolls would have a higher speed.

The table segments are the number of removable wear plates  attached to the table. For example, if there were 4 plates/segments, then the 4x table speed may produce vibrations. Similarly the 4 plates/segments passing 3 rolls would produce a frequency of 4x3= 12x table speed. Harmonics could be expected, especially if table surfaces are worn.

Have you measured vibration on mill including table support and roll support/springs?

Walt

Walt Strong posted:

What is the roll speed? From geometry the roll speed can be estimated by the ratio of table diameter (at point of roll contact) to roll diameter times table speed. I assume roll diameter is less than table rotation speed, so rolls would have a higher speed.

The table segments are the number of removable wear plates  attached to the table. For example, if there were 4 plates/segments, then the 4x table speed may produce vibrations. Similarly the 4 plates/segments passing 3 rolls would produce a frequency of 4x3= 12x table speed. Harmonics could be expected, especially if table surfaces are worn.

Have you measured vibration on mill including table support and roll support/springs?

Walt

please check both attachments tilting pads at one them and the other one is the gearbox section showing its internal components ,, i think you mean the tilting pads here by word segments am i right ?

Attachments

Walt Strong posted:

What is the roll speed? From geometry the roll speed can be estimated by the ratio of table diameter (at point of roll contact) to roll diameter times table speed. I assume roll diameter is less than table rotation speed, so rolls would have a higher speed.

The table segments are the number of removable wear plates  attached to the table. For example, if there were 4 plates/segments, then the 4x table speed may produce vibrations. Similarly the 4 plates/segments passing 3 rolls would produce a frequency of 4x3= 12x table speed. Harmonics could be expected, especially if table surfaces are worn.

Have you measured vibration on mill including table support and roll support/springs?

Walt

Here is the information I got 

Number of rollers 4

Segments 12

Diameter of table 4.6 m. /table speed 26.4 cpm

Diameter of rollers 2.3 m. 

 

 

 

 

Add Reply

Post
×
×
×
×