Question #7

Categories: Setup

What type of computer can we use with your equipment.
For testing on site, the ideal computer is a tablet PC pen computer (which we do not supply as part of the package).
We have excellent experience with those computers, and it is highly recommended.
You can use a regular notebook computer on site, but this is less convenient: entering data with the keyboard is difficult with dirty hands or when there is light rain. In addition, the color screen of notebook computers is almost impossible to read under direct sunlight. If you still wish to use your own computer for site work you may do so.
Ideal computer should meet those requirements:
1. Good sunlight visibility
2. Long battery life
3. Robust and waterproof

For office work you can use any Windows computer and printer.

Question #10

Categories: Setup

What operation systems can I install the software on?
Our software is tested on Windows XP, Vista (not recommended), Windows 7 and Windows 8
PET Bluetooth can also be used using Android OS

Question #11

Categories: Testing

What are the maximal length and diameter of testing piles?

PET can test of any diameter, while CHUM is limited by the necessity to install one or more access tubes in the pile during casting. Typically, cross hole testing is used for piles of 800 mm diameter or larger.

With the CHUM you can test piles of any diameter to depths of 140m.
With the PET there is a theoretical limit on the length-to-diameter (L/D) ratio, dictated by soil friction. It lies between 20 (hard soils) and 60 (soft soils).  

Question #12

Categories: Testing

Can we test integrity of reinforcement in piles and it length?

Question #13

Categories: Testing

How does the kind of concrete influence the test result?
Concrete piles are normally tested at an age of not less than 5 days. The concrete age and strength determine the wave velocity, and this has to be taken into account in the PET test.

Question #14

Categories: PET Testing

Can we test reinforced piles?
Does the reinforcement effect the test result?
Absolutely, you can test any concrete piles. The normal amount of reinforcement (<1%) has no influence on the results.

Question #15

Categories: CHUM Testing

It was observed that during the testing, we are losing 20 to 30cm of pile length, could you please explain.
This is the probe length.
Testing starts when the BOTTOM of the probe touches the BOTTOM of the pipe, but stops when the TOP of the probe reaches the TOP of the pipe.
If the pipe was only 31cm long - the probe can only move 1cm inside it.
This is OK and is inherent to this test method.

Question #16

Categories: Analyzing CHUM Reporting

We could see that in some cases, the Attenuation curve is touching the FAT curve and in some cases the Attenuation curves come first and then the FAT curve and Vice versa also. Has this got any implications?
No! FAT and attenuation have different units.
The curves are plotted on the same area in order to save space and nothing more. If I was to pick different axes values, the curves would touch in different, arbitrary point.

Question #17

Categories: CHUM Testing

Does pulling velocity has any effect on the test result?
Pulling speed has no effect on the results. The encoder constantly transmits the depth to the instrument, and a pulse is sent in predetermined vertical spacing. 

Question #18

Categories: CHUM Testing

Our CHUM is not showing the exact spacing between the sonic tubes. (for eg. When we measure 0.64m in the site the CHUM is showing 0.88m).we have tried this out in some 10 piles and still the error is existing.
In the leveling screen, the distance is only an estimated (this is why it has the ~ sign next to it) CHUM can only measure time, not distance! The distance is calculated assuming a signal velocity of 4300m/s. In your case, the signal velocity might be lower - the concrete might be too fresh or of lower grade/quality. Anyway - the distance value is presented in this screen in order to help you level the pulses, not for reporting.

Question #19

Categories: Analyzing CHUM

What is the criteria for FAT & Attenuation to judge pile integrity i.e. during our testing, some FAT will be 184µs and for other pile it will go to 260µs. Also how FAT is related to concrete quality & what is the range of values of FAT &Attenuation for good & bad piles. Please advise with examples.
FATstands for First Arrival Time. Since the wave velocity in cured concrete is about 4300m/s ( /- 10%), it will take the wave 1ms to pass 4.3m, 500us to pass 2.15m and 250us to pass 1.075m etc. Since sections are never at the exact same distance, it is expected that the FAT will change. 184us is typical for a section of 0.8m and 260us corresponds to 1.1m If your FAT is too high for the section distance, the concrete might not be cured or it is of low quality/grade.
The received signal has a measurable energy. The energy drops exponentially as the distance grows. It is expected for the energy to remain roughly constant in a homogeneous concrete. Attenuation is simply the energy in decibels (DB) units, relative to the maximal energy that can be recorded. A change of 6db means half the energy. 12db is 1/4 the energy, etc. 

Question #20

Categories: Analyzing CHUM

A number of the traces show a very large drop in the relative energy, but little or no change in the FAT.  I understand that it is generally taken that the FAT gives the biggest indication of inclusions or defects, however, I note on your web site that a low relative energy means a "weak" signal.  When would I expect to get a weak signal, and could it represent an anomaly even if the FAT`s are consistent?
weaker signal with no change in FAT can be generated when the wave path becomes smaller, for example, in the case of a necking.

As seen in this image, the shortest path between source and receiver is not disturbed, therefore the FAT shows no indication of the defect. However, the total amount of energy reaching the receiver is significantly reduced. This is just one way of explaining this phenomenon. I am not aware of any research done to quantify it or find additional explanations. There is no way to assess the size and severity of the defect (if it is indeed a defect) from this trace, but you quite confidently assess that if the energy dropped significantly (say by 18 db or more in the attenuation curve) you are not crying wolf.

Question #21

Categories: Analyzing CHUM

In general the traces are quite smooth (about a vertical line), but every now and then the line is very jagged, although still about a vertical line. Does this sound like a problem?
Concrete is not a homogeneous material and some variations of the FAT and attenuation are to be expected even in the most controlled environment. Unless you can see a significant change in FAT or energy (And I cannot quantify this change) you are safe.

Question #22

Categories: Analyzing CHUM

How can we estimate the distorted area of the pile, from FAT & Attenuation?
There is no standard, and you have to apply solid judgment for changes in FAT or attenuation. I recommend that you will send the files to us. We will be glad to assist you with the interpretation until you feel confident enough.

Question #23

Categories: Analyzing CHUM

Which is the easiest method of evaluating a pile, like FAT vs. Attenuation or FAT vs. Energy. Explain.
Those are equivalent.
Originally CHUM only showed Energy (without units). We introduced Attenuation in a later version, but kept the energy option for backward compatibility with those users who already got used to present and submit energy curves in their reports.

Question #24

Categories: Analyzing CHUM Reporting

In our Equipment, for the PRESENTATION of results, we are allowed to choose from FAT, Attenuation, Relative Energy & Apparent Velocity. BUT, when we choose the Apparent Velocity option, We are able to see a green legend for Apparent Velocity in the bottom of the graph & also the velocity written in m/s., BUT there is no representation of it in the graph.
CHUM can only measure arrival time, not signal velocity. In order to plot the velocity curve, you must enter the horizontal spacing between the pipes. Once you enter this value, you will be able to see the velocity curve.

Question #26

Categories: CHUM Testing Theory

What are the advantages and disadvantages of CSL compared to the radioactive test method?
here is a comparison table






Compressive strength

Max. range

50 -- 75 mm

4-5 m

Min. concrete age

A few hours

~ 5 days


<100 m/hr.

700 m/hr.



ASTM D6760-08

Worldwide operators

Very few

Many hundreds



Both 2D and 3D

Air transport


No limitation

Dangerous materials regulation



Current R&D

Practically none




Practically all

Need of a calibration pile





Question #27

Categories: PET

Can the PET be used to determine the concrete strength?
Yes, it can. If the length of the pile is known, the PET can be used to determine the wave speed in the pile.
The wave speed c  is related to the compressive strength f by the following equation
       c = K f1/6

Question #28

Categories: PET

Can the PET results serve to find the length of the pile and its profile (bulging, necking?)
Yes. The PET can provide the pile length to an accuracy of 5 to 10 percent. Changes in the cross section can be found by visual inspection of the reflectograms or by signal-matching that is built in the PET software.

Question #29

Categories: PET

Where can I find a spare waterproof cable for the USB PET?

We make a high-quality rugged polyurethane, right-angled cable  - inquire at

A compatible (lower-cost) cable is available commercially:
The USB cable is a standard cable made by Bulgin, UK.
Part number is PX0441
Smaller quantities can be found in specialized retailers such as (Worldwide branches)

Question #30

Categories: PET

Where can I get more putty for the PET?
The putty is made by HBM, Germany
Part number is AK22
We also offer this directly.

Question #31

Categories: PET

Where can I find spare hammer heads and hammers for the PET?
Hammers and hammer heads are manufactured by Thor (UK)
Part numbers:
  • Hammer: 11-710 
  • Hammer heads: 76-710NF

Question #32

Categories: CHUM PET PSI Testing

How can I test a pile that already has a pile cap on?
The following methods may apply:

1. Parallel seismic
Requires a small parallel access tube to be installed very close to the existing pile. It can only give information regarding the length of the pile, nothing more. the advantage is that it does not require any access to the top of the pile and you can use the superstructure (the pile cap in this case) to generate an impulse.

2. Pulse-echo
Can provide much more information about the pile, including length, major features/defects, etc. the test is also very quick and require almost no preparation. However, you must have access to the top of the pile.
If you can find or cut with a disc-saw a very small groove (about 3cm deep) close to the top of the pile you can do the test from the side of the pile. see image.
In this case you will have the downwards waves mixed with the reflections with the superstructure. this makes the success of the test less predictable.
However, you could test all piles with relatively small effort and gain a lot of information.

3. Single-hole ultrasonic testing
An access hole must be drilled throughout the length of the pile (which is tricky) and the hole used for SHUT
In case a defect is found, the hole can use them for reinforcing the piles by grouting a Dywidag bar in the hole.

Question #33


I cannot save a pile on Windows 7/8, should I upgrade the software?

The most likely reason is that the [Home folder] is set to a folder which requires administrator privileges to write on, such as \program files, or c:\

There is an easy way to check this: Run the software as an administrator (right click the software icon and select [Run as administrator])

If this enables you to save, all you should do is move your home folder to a suitable location, such as your documents folder, or the public documents folder.

For more information, consult the user manual regarding projects and home folder.

Question #34


Which Windows version does CHUM support? 
What about 64bit support?
Yes CHUM software runs well on WinXP/7/8/10 and on 64 bit systems
Piletest device drivers are approved and signed by Microsoft

If you have a CD with old drivers - (dated before 2014) - please first login to the user community area and download the latest signed drivers

Follow those steps:
  • start device manager [Start]-(Right click)[Computer]-[Properties]-[Device manager]
  • plug in the CHUM
  • locate the CHUM under [Universal serial bus controllers]
  • right click the CHUM icon and select [Update driver software]
  • Instruct the [new hardware found] wizard to locate the drivers on your hard drive on [\program files\\USB driver]
Contact if this did not solve the issue

Question #35

Categories: PET Setup

Which Windows version does PET support? 
What about 64 bits?

Yes PET software runs well and tested on Win XP/7/8/10 and on 64 bit systems
Piletest device drivers are signed by Microsoft

If you have a CD with old drivers - (dated before 2014) - please first login to the user community area and download the latest signed drivers

If you are having troubles installing the drivers, Follow those steps:
  • start device manager [Start]-(Right click)[Computer]-[Properties]-[Device manager]
  • plug in the PET
  • locate the PET under [Universal serial bus controllers]
  • right click the PET icon and select [Update driver software]
  • Instruct the [new hardware found] wizard to locate the drivers on your hard drive on [\program files\\USB driver]
Contact if this did not solve the issue

Question #36


Why our PET signal graph does not has unit for the Y axis?

In principle, the Y axis is the head velocity in m/s
However, since the pulse goes intensive signal processing, including filtering, sharpening, rotating, shifting and exponential amplification, the Y axis can no longer be used quantitatively and no scale is provided.
The units on the vertical axis have some meaning only at the head or close to it, where the amplification is basically 1 (no amplification). Adding velocity units may therefore be grossly misleading.
It would have been possible to draw equal-velocity lines on the graph - but this would add clutter and no information.
PET test is qualitative - you should get the same result if the impact is strong or weak (within reasonable limits, of course) .

Question #38

Categories: PET Theory

Why doesn^'t  PET software calculate the "Beta value" for a pile?

The so-called Beta Method (Rausche and Goble 1979) was originally proposed for assessing damage during pile driving. 

Even though its validity was never independently assessed (as the authors themselves (note in their paper), it gained some popularity since it determines automatically the amount of pile damage and frees the engineer from the burden of thinking. 

In any case, it was never meant to be used on bored piles and with time, it became clear that this method is meaningless even in driven piles.  

In view of the circumstances, decided long ago not to support it.

If any competitors produces such beta values they are misleading their clients and themselves, too. 


ref: Rausche, F. and Goble, G.G., "Determination of Pile Damage by Top Measurements," Behavior of Deep Foundations, ASTM STP 670, Raymond Lundgren, Ed., ASTM, 1979, pp 500-506.

Question #39

Categories: Analyzing CHUM Theory

Why is the apparent velocity higher than what I expected?

The point is understanding the ULTRASONIC METHOD
Sound waves are not like laser beams - they can travel in curved path and will go around defects!
When the wave has to pass a defect, the travel time gets longer, but it needs to be quite a large defect (or very close to the tube) to create a significant change in arrival time.
On the other hand - the defects does block the energy and you will see a significant energy reduction
Here is a drawing and explanation

You can see that although the defect is quite large, not much length is needed to go around it.

However, a significant part of the wave-front if blocked and the total energy arriving to the receiver is reduced

If we will make a simplification of the wave path - using L and r - it becomes very simple to calculate - and good enough approximation

Here is a table with typical values - defect size and velocity change - you can see that for 1m spacing the defect has to be ~60cm in diamater (r=0.3) to see a 20% change in arrival time/velocity.

Extra path
1.00 0.00 0.0% 4400
1.00 0.05 0.5% 4378
1.00 0.10 2.0% 4315
1.00 0.15 4.4% 4214
1.00 0.20 7.7% 4085
1.00 0.25 11.8% 3935
1.00 0.30 16.6% 3773
1.00 0.35 22.1% 3605
1.00 0.40 28.1% 3436
1.00 0.45 34.5% 3270
1.00 0.50 41.4% 3111


  1. Small changes in arrival time can point quite large defects - depending on defect location and size
  2. Energy is a very important part of the evaluation

Question #40

Categories: PET Testing

Can I test piles that are under a pile cap by sonic method?

The answer is: it depends...

See here a good example

Credit: CFT & Asociados, S.L.‎

Question #41

Categories: CHUM Reporting

How do I do classification of test pile condition at Satisfactory, Anomaly, Flaw, and Defect

Those classifications terms are wrong
Anomaly is not a small flaw and flaw is not a small defect - those are different things - not different severity levels!

here are the definitions we recommend

Any irregular feature in the NDT results. 
An anomaly may be due to the:
  • Testing instrument (such as noise)
  • Means used (access tube de-bonding)
  • Surrounding soil (abrupt changes of soil friction) 
  • Pile (Actual flaw or defect)
  Any deviation from the planned shape or composition of the pile, which does not necessarily detract from the pile^^^'s performance. 
A flaw that, because of either size or location, may affect the pile^^'s capacity or durability. 

The Test Laboratory is responsible for reporting anomalies and for assessing which anomaly comprises a flaw
The Geotechnical Engineer and the Structural Engineer are jointly responsible for the decision which flaw comprises a defect.

According to our definitions, you find anomalies, and use your best judgement to decide if those are actually flaws, or just setup issues
Once you decide it^^^'s a flaw, you can decide on "minor flaw"..."critical flaw" - but it is up to the DESIGNING ENGINERR to decide if this flawed pile is defective (rejected) or accepted as-is, etc.

There is no one simple number to decide on this - you have to decide case by case, collect ALL the supporting information, setup method, site investigation, ask questions and reach a conclusion

Any classification system that is just based on one number will make very bad FALSE calls. Those mistakes can be very expensive.
There are no shortcuts, you have to apply judgement on each case.

Question #42

Categories: Analyzing CHUM Theory

In CHUM 3D Tomography, how come I see low velocity zones outside the cage, even when I use just straight-ray calculations (Draft)?
The basic fact is that waves travel in curved lines, and thus the CSL method is effected by the cross-section area outside of the polygon formed by the access tubes.
But why is this showing in the 3DT?
When doing curved-ray calculations (anything but "draft" mode in CHUM3DT), this is understandable. the software iteratively is trying to find the fastest wave path, which can easily be outside the above polygon. the software is trying to simulate a situation which is as close as possible to what is happening in real life.

But why is this happening also is straight-ray calculations?...
After all, every voxel is an unknown parameter in the big matrix, and each pulse is exactly one linear equation which is formed by passing a straight line between the two tubes. Voxels which are outside the straight path do not appear in ANY equation and are therefore totally undetermined... We can not see why a simple matrix inversion calculation will assign any value to those voxels...
A possible explanation is: it^^'s not the math it^^'s the graphics...
The voxels are quite large, typically 20cm^3 (8 inch cubes). they form a coarse grid, and each eventually gets a velocity value.
This already can cause some projection outside the "polygon" because voxels and tubes are not necessarily aligned.
After the calculation is completed, the software traces contour lines. contours are by nature an interpolation which can show as projecting even further out from the polygon.
So the graphics is misleading... contours makes it look nice and smooth, but the actual calculation granularity is quite coarse...
See picture below, the pile is 1.6m, each voxel is 20x20x20, 5 access tubes... gray voxels are either good" - or not touched by any straight path. I defined profile 1-5 as low-velocity (red).
To draw the contours I used free-hand, assuming that the velocity point is at the center of each voxel.
I hope you agree this is a simple and reasonable explanation

Question #43

Categories: Analyzing CHUM

Is there any recommendation or guideline for using "Filter" during the test?
We found the results might different if we use different level of "Filter". 
Under most conditions the FAT line shows added random noise - mainly due to poor FAT picking. This can be remedied by manual FAT picking and/or by filtering. 
As a rule, filtering hides data and thus has to be used as little as possible - just to hide the random noise. Excessive filtering also hides legitimate data.
If a filter setting of 1 or 2 does not do the job, it^^'s good practice to repeat the test with careful pulling or repeat the test after waiting  a few days.

Question #44

Categories: CHUM

How can I measure CHUM sensor frequency?
To measure the sensor frequency, open the FAT utility with a typical strong pulse, 
Point the cursor on an easily identified peak and note the time
Count 10 peaks to the right, click again and note the time
the frequency is 1/(t2-t1) 
Since the utility gives the time in microseconds, you measure 10 cycles and you want the result in KHz, the calculation is 10000/(t2-t1)
see image below

Question #45

Categories: Analyzing PET Theory

Does PET Software draw Pile impedance Profile (shape of pile)?

Yes - it does. In the PET main window select the pile you wish to draw,  press [Pile]-[Match]. 

See the user manual how to use this feature

To solve the problem automatically Choose the Match Range and press the [Match]  button at the bottom left. At any time you may stop to change the friction profile and/or the impedance profile and then resume the matching process until a good fit is obtained.

Important Note:

Inferring the pile^'s shape from the reflectogram is an "inverse problem" (since the shape is the input and the reflectogram is the output)

Many Inverse problems, including this one are not unique - meaning that there are many combinations of pile shapes and friction profiles that can generate a similar reflectogram, hence no algorithm can plot the pile shape, just A pile shape, one of many possible.

However, sometimes all the possible solutions explain features in the reflectogram in a very similar way (e.g. they all show a nacking at a certain depth)

Piletest recommends using the signal matching feature as a DECISION SUPPORTING TOOL and not as a magic wand that saves the need for any thought and good old engineering judgement. 


Inferring Pile Shape from Pulse-Echo Test Records by Evolutionary Algorithm

Amir, E.I & Amir, J.M

The 9th International Conference on Testing and Design Methods for Deep Foundations

IS-Kanazawa 2012


Question #46


What are the parts of the CHUM software?
The main CHUM software (CHUM.EXE) is used for 
  • connecting with the CHUM hardware
  • taking the measurements in the field
  • Analysis 
  • for producing reports 
This is 99% of the work - beginners should only use CHUM.EXE

The other parts are needed ONLY for 3D tomography (3DT):
Our 3DT solution has 2 parts
  1. The 3DT viewer (CHUM3DTViewer.exe)  - this part displays the 3DT data - lets you rotate, zoom, etc and create a report - this part is free to distribute
    The viewer can be started from the main CHUM software
  2. The 3DT engine  (GEOTOMCINI.EXE) - this part has NO user interface at all. it calculates the tomography. this part of the solution is not free ($3300) and has a limited distribution license
    The engine is started by the viewer only and runs in the background

Even the most basic CHUM system has all the hardware and software to COLLECT data for 2DT and 3DT

Users who want to experiment can send us the files and we do the engine calculations for them - for a limited time/amount
Users who have the engine can send the results and the viewer to anyone who wants to view the results on the software, or they can produce a standard document report
The engine must
not be distributed

 Mailing address Limited
18 Fouracres Walk
Hemel Hempstead
Herts HP3 9LB
United Kingdom