In the September issue (page 38) of Australian Welding, Weld Australia highlighted the progress Xiris has made in engineering high-dynamic-range (HDR) cameras that improve weld quality and efficiency.

We’re proud and gratified that Weld Australia recognized us. The organization has been a tireless proponent of the importance of welding, spreading the word that many industries wouldn’t be the same (or even exist) without welding.

All companies in those industries can benefit by using HDR cameras to monitor welding processes in real-time. Visual monitoring with eyes simply can’t match the clarity and certainty of a video record of the welding as it’s happening.

Australian Welding explains: “It is an enabling technology used across a wide range of industries and applications, from micro-joining of medical devices, electronics, and photonics to larger-scale applications such as mining equipment, pressure vessels, ships, rail transport, water pipelines, and components.

 “Recent advances in advanced CMOS image-sensor design and supporting electronics allow best­in-class cameras to provide high-contrast images without  saturation of the brightest portion of the image, exposing the details of the electrode, melt pool, seam, and surroundings.”

The article also details how remote camera monitoring benefits employees’ health and safety. Using weld cameras helps operators ergonomically because they’re not continually contouring their bodies to see the weld. They are also safer because they don’t have to be near the flame.

Another plus:  videos of in-process welding can be highly useful in training and continuous improvement.  It’s a technology that will attract and retain welding apprentices.

You can read the full article here.

With the Xiris View Cameras, welding videos can be displayed live, and simultaneously recorded in raw or .avi format. By having a digital video of the weld process, operators are able to monitor the weld process live and make adjustments on the fly, or review the weld process offline for quality or audit processes. Sometimes, a standard view of the original video does not provide the operator with enough detail to see subtle features in the weld process that may be of interest. Dross forming in the melt pool, the solidification pattern of the cooling weld bead, or improper melting of wire or powder being fed to the process are all examples of this. In certain cases, it is interesting to apply a convolution tool to accentuate subtle greyscale differences to make some features more visible to the operator.

The Xiris Convolution Tool

Convolution is a simple mathematical operation that provides a way of `multiplying together' two arrays of numbers, typically an image with a kernel of numbers, to produce a revised image of the same size as the original image. It is a form of image processing that uses a mathematical operator to create an output image that is a linear combination of an original input image.

The convolution is performed by applying the kernel at each position of the input image and generating a new pixel at that location by mathematically combining the elements in the kernel with the neighboring pixels in the image. The result can strengthen the visibility of edges or other features that are hard to see in the original video.

Fig.1 Original Image of Melt Pool		Fig. 2 Laplacian Convolution with 3 x 3 kernel

As seen above in Fig.2, a simple Laplacian convolution (a type of convolution that is very sensitive to noise) can expose some subtle details that are difficult to see in the melt pool area of the original image.

While applying a convolution tool may not be suitable for all applications of Xiris weld cameras, it can certainly expose some subtle details in the images that are hard to see in the original, raw video of a welding process and has the ability to detect special features of interest in certain situations.

Highly precise welding procedures commonly found in power generation applications where thick walled sections of pipe are welded together using multi-pass TIG welding procedures have demanding quality monitoring requirements. One of the most important monitoring tasks is to monitor the quality of the melt pool from the back side of the weld. In the case of orbital welding two sections of pipe together, this is often done by using a Xiris weld camera mounted inside the pipe looking at the backside of the root pass of an orbital welding head operating from the outside of the pipe.

If a Xiris weld camera is mounted inside the pipe looking at the melt pool, what is visible is a high intensity spot surrounded by an area of lower intensity (see Figure 1 below).  Precision fabricators are often interested in the shape, size and location of the precise boundary of the melt pool relative to the weld seam. Segmentation of the edge of the melt pool is therefore an important first step to measure the shape of the melt pool.

Xiris was given this task to first detect the edges of the melt pool, then measure the size, various shape parameters, and the center of the object relative to the weld seam.

Fig. 1:  Original Image of Melt Pool		Fig. 2 Segmented Region of Interest

Once the Melt Pool can be correctly segmented from the rest of the image, the Xiris Blob Analysis tool can be used to compute various measurements of the region of interest, such as; centroid, area, perimeter, bounding box, and roundness, that together, provide information such as; to how well the melt pool is forming, if the welding torch is burning through, if the weld torch is tracking the seam properly and other process features.

Image processing tools, like Blob Analysis, can be used to find valuable shape, size and location information of key features of the welding process. When combined with a Xiris weld camera, real time analysis can be achieved.

Follow Xiris on social media for regular updates and welding videos!

At Xiris we believe in giving back to the communities in which we live and work by creating a positive impact to support our surrounding neighbourhoods. As we enter the new year, we wanted to reflect back on all the initiatives our dedicated team took part in during 2019.

Burlington Food Drive

As part of our commitment to the community, we participated in the Burlington Food Bank food drive during the months surrounding the holiday season. The Burlington Food Bank ensures that no one in Burlington struggles with hunger and we wanted to do our part to ensure that every family could enjoy the holidays. In October our team collected and donated 83 kg of non-perishable food items and over 45 kg in December.

Skip welding, otherwise known as Tack Welding, is a commonly used welding technique used by fabricators to help limit excess heat penetration in, and therefore minimize distortion of the parent material during welding. However, sometimes weld skips can occur unexpectedly and disrupt the process of a good weld. Weld skips can be difficult to predict because they are random events that leave weld operators scratching their heads for a cause.

Fig. 1 Weld Skip on a Section of Welded Tube

An Interesting find!

Recently, at a customer’s site, a robot was performing a longitudinal weld on a large component used in the agricultural industry when a random weld defect was identified by an employee monitoring the weld process.  Unknowingly, it turned out that approximately one in every twenty welds were made with this defect.   Not knowing how to correct the defect, the customer invested in a Xiris weld camera.  After setting up the weld camera to get a better look at the welding live, the weld camera revealed that the weld torch would suddenly skip over a segment of the weld seam that was supposed to be welded, something that is commonly referred to in the welding industry as a “weld skip”. But the question still remained: what was causing the weld to skip?

After carefully examining the recorded video from the weld process, it became clear to the welding engineers what was triggering the defect: an interference in the weld had caused the weld torch to momentarily become stuck in position before the pulling force would build up to enough to cause it to come loose.  Once the torch would come loose, it would jump over a segment of the seam before resuming the weld, resulting in a weld skip.

How Xiris Weld Cameras Can Help

Weld skips can be disastrous to a production process, causing product defects which can be costly to rework.  Having an employee continuously monitor a weld process can be unsafe and an unproductive use of their time, resulting in lost time, money and production. As a result, for some fabricators, their weld processes need to be constantly monitored for their weld skip. This is where Xiris’ weld cameras can help by providing real-time, close up views of the weld process while it is happening, capturing live images of potential welding defects for the operator who can make process parameter changes before the defects turns in to a major production problem. The Xiris weld cameras have the ability to archive the weld videos to allow operators to review the videos offline and analyze what caused a weld defect to occur.  In addition, using a Weld camera can provide employees with a safe means of remotely viewing the welding process, avoiding the need to monitor the welding process in hazardous or confined work spaces, increasing employee focus on other operations.

Experienced, professional welders agree that sound is often just as valuable in monitoring welding as being able to see what is going on in the welding process. Issues with wire feed speed, voltage, shielding gas, current and other parameter inputs can all be heard, sometimes more than they can be seen. Giving operators another tool to improve their welding operations and the quality of their products is why Xiris has added audio capture to its industry-leading weld cameras.

Today, we’re sharing 3 ways audio can help your welding process.

1. Multisensory monitoring improves quality. There are many ways to monitor welding operations, such as current, voltage, visual inspection, machine vision sensing, thermal imaging and more. Each method has its own advantages as well as shortcomings. Using multiple monitoring techniques can minimize, or even eliminate these shortcomings.

Audio is another tool to add to your toolbox. Acoustic monitoring may help to overcome some of the challenges of monitoring “dynamic welding characteristics, especially near the arc due to the high temperature, spatter formation, fumes, etc.” ¹

Many factors can affect the generation of welding sounds, such as the volume of the burning arc, the shielding gas flowrate, vibration of the droplet transfer into the welded pool and the noise of the welding apparatus.

Xiris’ weld cameras, equipped with high dynamic range imaging and now equipped with audio monitoring, can help operators be confident in the quality of their welds and decrease the time spent on post-weld inspection and rework.

For over 10 years, Xiris has focused on helping our customers to detect, identify, address and prevent defects in their welding processes.  We know that better images lead to better decisions and better process control. Now, customers can combine those images with audio to further improve their decision-making process.

¹ Lv N., Chen S. (2011) Investigation on Acoustic Signals for On-line Monitoring of Welding. In: Tarn TJ., Chen SB., Fang G. (eds) Robotic Welding, Intelligence and Automation. Lecture Notes in Electrical Engineering, vol 88. Springer, Berlin, Heidelberg

Follow Xiris on social media for regular updates and welding videos!