X-Ray Forum for NDT Users
It is currently 17.06.2021, 01:02

All times are UTC + 1 hour [ DST ]




Post new topic Reply to topic  [ 2 posts ] 
Author Message
PostPosted: 15.01.2021, 15:44 
Offline
User avatar

Joined: 19.10.2019, 16:47
Posts: 115
There are a few different materials for Image Quality Indicators (IQIs). The reason is, that the IQIs should fit to the absorption of the material in test. E.g. in ASTM E1025 nine different material groups are defined increases in attenuation, from Material Group 001 for polymer matrix composite materials, over 03 for magnesium, 02 aluminum, 01 titanium to the more heavy material groups without the Zero in the beginning. Examples are the Material Group 1 for stainless steel, Material Group 3 for Inconel, and Material Group 5 for phosphor bronze like gun metal.
As long as the IQI is only a percentage of the total material in the beam, the quality of the material is not of highest interest. But some standards as ASTM E2597 or E2737 use IQIs which are totally exposed and all material imperfections will be visible and may influence the results. As detectors - especially DDAs - improved a lot in quality, these imperfections of the "base material" are on a level that they dominate the results. An example could be seen where the material imperfections of the five groove wedge dominates the visibility of the grooves at longer exposure times.

Therefore a task group at ASTM was found to evaluate materials for light metal (e.g. Aluminum) and heavy metal (e.g. steel) which have high homogeneity or uniformity and low internal structures. The topic for Aluminum is handled .
The first image shows the big difference of 10mm steel plates where the bad material limits the achievable SNR to about 325 where as a better material is good for a SNR of about 1700 with the same exposure conditions:

A simple test with different IQIs should show the difference, the plate hole IQI of ASTM E1742 for an EPS value of 0.78 for the 1T hole of the .25 IQI and the wire sets where the smallest wire is 50µm thick.

With 64s exposure time at 200kV, 6mA, 1100mm FDD the difference is very well visible.

Reducing the exposure time to 4s will still show a big difference:

and it is quite funny that the 1T hole of the 0.25 IQI is quite good visible – or is it another indication? It is hard to separate between the IQI hole and a “natural imperfection” of the material. Here the wire sets offer an advantage as the wire shall be clear visible along 10mm in minimum reducing the risk to confound with a material imperfection.
Next picture shows the setup for the test in the next post:


Top
 Profile  
Reply with quote  
PostPosted: 15.01.2021, 16:19 
Offline
User avatar

Joined: 19.10.2019, 16:47
Posts: 115
The following materials were in the test:
Image
Additional to the Inconel plate – which is required in E2597 - we also took a plate of brass as reference because brass is known to have nearly no internal structures and should be the reference.
Within the test we took images of each material with 1s, 4s, 16s, and 64s exposure time. For the presentation of the results here the 64s values (longest time) are used.
Of each material we took two images at identical positions; the subtraction of both images shows only the X-ray noise as all structures are removed by the subtraction.
Then we moved the plate by about 10mm in both directions (0.7“diagonal) and took a third image of each material with the same parameters as before. The subtraction of third images from the first image shows the fixed pattern noise (material structure) and the X-ray noise. As we have the X-ray noise from the previous image series we can subtract the X-ray noise (variances) from the total noise and get the fixed pattern noise.
The first picture shows the signal level of the 10mm plates with 200kV.

The highest absorption comes with brass, Inconel is close. Most of the steels have nearly double the signal level, only SS316L (SS1.4404) has a higher absorption.

The total SNR (SNR of X-Ray and Material) shows the achievable performance level which you could measure in practice.

From the difference images we get the SNR of X-Ray which is proportional to the square root of the signal. This is conforming to the theory.
Now we can calculate the fixed pattern noise of the material which I call the SNR of the material.

Here we see very big differences for the different steels. Also the Inconel performs quite good – the Inconel for the test got heat treatment to remove most of the structures, which was quite successful. But nevertheless the steel SS316L (SS 1.4404) performs 60% better …
Finally also the visual impression is important to accept a new material:

Brass is very homogeneous and show only little structures; Steel 316L (SS 1.4404) show moderate structures with wide spread; this heat treated Inconel performs quite well, but the structures in medium size are well visible and should be avoided in an IQI.
In the relationship to the X-Ray noise we see a clear winner: Brass.

As the workgroup in ASTM decided not to use brass or cooper, the recommendation for Steel 316L is obvious:
- it shows fewer structures than Inconel
- it provides a better SNR than Inconel or other steels
- it does not need a special treatment as Inconel does
- it is more easy to machine
- it is cheaper and more easy available on the international market.
Finally it has the highest absorption of all steel alloys in the test and is therefore closest to the Inconel.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 2 posts ] 

All times are UTC + 1 hour [ DST ]


Who is online

No registered users


Similar topics

Material for IQIs - Aluminum (light metal)
Forum: IQI (Image Quality Indicator)
Author: Klaus
Replies: 1

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
cron



Hosted by iphpbb3.com

Impressum | Datenschutz