Electron beam probing systems - a review

TWI Bulletin, May/June 2001

Olivier Nello joined TWI in 1997 after completing his National Service in Africa as tutor in Electronics at the Ecole Africaine de la Meteorologie et de l'Aviation Civile. He gained an MSc in Electronics at the Pierre et Marie Curie University in Paris and is currently studying for an MBA at the Open University Business School. He has worked in the Electron Beam and Friction and Forge Processes department for four years, mainly involved in equipment development and research such as beam deflection and beam characterisation.

In an increasingly competitive environment, novel welding techniques for aerospace structures are being developed with the aims of cost reduction and improved quality. In electron beam (EB) welding, as Olivier Nello reports, the use of beam probing techniques enables aerospace manufacturers to achieve these objectives.

Many electron beam (EB) welding systems are used for precision work in which the reproducibility of the beam is paramount. Unfortunately very small changes in the gun or machine parameters can lead to large changes in the beam profile and erratic welding performance. Lack of reproducibility, such as in the focal position of the EB or the beam intensity, are the most common problems reported to TWI from industry.

Various techniques exist for ensuring beam quality, from welding of testpieces to direct measurement of the beam itself. Although testpiece welding can confirm good beam quality the process is often time-consuming and the results, if not acceptable, are difficult to interpret in terms of machine adjustment. Beam probing techniques potentially offer both an assurance of beam reproducibility and information that could be more readily interpreted.

Edge detection probing system

The signal obtained can be mathematically processed to provide approximate energy density data. However, this type of system is best used for an approximate determination of focus position.

The system can be simple and robust mechanically, but the signal obtained is prone to interference from electrical noise. Loss of elastically back-scattered electrons from the detector, secondary electron emission and the generation of positive ions can cause degradation of the signal.

In addition, if the detector edge becomes badly eroded, the signal deteriorates. If a first-order differential is applied to the signal obtained, the data is similar in form to that obtained by slit or rotary wire probing systems and is then far easier to interpret. No information about beam ovality is given unless two or more signals are obtained using detectors oriented at different angles.

Wire probing system

Wire probing systems use a small diameter wire to sample the beam. Again the passage of the wire through the beam is achieved either electrically or mechanically. The resolution of the system depends entirely on the wire diameter being as small as possible.

The minimum wire diameter depends on the wire material, the probing duty cycle, and the beam intensity. By placing more than one wire in the path of the beam, several measurements at different work heights can be made at approximately the same time.

The system suffers from similar difficulties with electrical noise as edge detection probing systems. The interference from positive ions emitted by the main beam target can be alleviated by capturing the beam in a cup with a small diameter entrance hole. Positive ions can also be generated from the wire, especially when using small diameter wires. A further disadvantage is that there is some risk of contamination, caused by wire debris, of the workpiece in the chamber.

The practical minimum wire diameter is 0.1-0.2mm, which limits the resolution of the system. Rotary probes use a delicate rotating arm, which requires careful balancing before use. The data obtained is readily interpreted, but separate data collection or multiple units must be used to obtain information regarding beam spot ovality.

Slit probing system

Slit probing systems ( Fig.1) consist of a refractory metal slit through which a small portion of the beam current passes as the beam is deflected electrically (at ~1km/sec) over the slit. The electrons passing through the slit are captured in a Faraday cup and a signal similar to a differentiated edge probe or wire probe signal is obtained.

The detector element itself is fully enclosed, and only subjected to a tiny portion of the beam power. This means that the system is quite immune to electrical noise pickup and signal degradation due to ion emission, which can degrade edge and wire probe signals significantly.

A typical slit width would be 0.02-0.1mm, which gives a significantly improved resolution compared with many other systems. Slit probing systems can be used to obtain information about beam spot ovality, focal position of the beam and even beam convergence angle.

Pinhole probing system

This technique can be beneficial in that it provides a complete 3D map of the beam energy density distribution. However, precise control of the beam deflection is required for the multiple data traces to be accurately overlaid, which can be difficult. The pinhole probe itself can be more problematic than an equivalent slit probe because of the need to drill and maintain a tiny hole in the detector; debris can easily block the hole, leading to loss of signal.

Data capture and processing is not as swift as with less sophisticated probing systems, simply because there is a lot more data to capture and process; also a certain amount of additional processing must be done in order to make the data useful.

Strategy for data collection and analysis

Beam ripple can occur on the beam current (originating from the auxiliary power supplies or the high voltage power supply) and on the accelerating voltage. Significant beam ripples are usually in the range 10 ¹-10 ⁴Hz and may cause changes in beam current and shape.

As the beam passes over a slit, wire or pinhole, a snapshot of the beam energy density distribution is taken. With slit or wire probing systems, the beam may be averaged over several scans in order to compensate for any beam ripple or alternatively the scan rate can be synchronised with a selected ripple frequency.

However, with a pinhole system, there is something of a dilemma. If the many scans required are executed over a period of 1-10msec, then the beam could go through a significant portion of its ripple cycle. The resulting 3D energy density distribution image could therefore be quite misleading.

In addition, there is another difficulty. The separate data traces must be accurately overlaid to give a true picture of the beam. With separate sweeps as described above, the timing of the beam over the pinhole may vary significantly with changes in the accelerating voltage. The beam deflection will also vary across several sweeps leading to a superpositional error of the separate data traces.

Whichever probing system is adopted, the fact remains that, if the beam ripple is significant then care must be taken in interpreting the results and it may be necessary to adopt a number of measurements of the EB in order to determine the energy density distribution throughout the ripple cycle.

TWI slit probing system

As part of a European funded programme aimed at improving manufacturing techniques in the aerospace industry, TWI has developed a portable slit probing system ( Fig.3). This system has been designed to be fitted to any EB welding machine that could accept the appropriate feedthrough.

The probe is a rugged device that can be left in the chamber permanently or moved from one machine to another to allow comparison. TWI slit probing system is capable of a wide variety of beam diagnostic tasks, including:

• Estimation of:

  - focus

  - beam profile

  - beam asymmetry

  - beam current

  - beam diameter

  
  
• Comparison of:

  - probe traces

Estimation of focus

With the slit probing system an accurate measurement of sharp focus can be obtained without applying a power beam to the workpiece itself. Simply by observing the various energy density distributions of EB, the machine operator will be able to unequivocally determine the focus setting, which gives the narrowest and highest signal peak at a particular beam current.

Estimation of beam profile

By scanning the EB over a slit parallel to the welding direction the transverse elevation of the energy density distribution of the beam may be determined. It is this characteristic distribution which is primarily responsible for giving an EB weld its characteristic shape.

Estimation of beam asymmetry

By scanning the EB over slits both parallel and orthogonal to the welding direction, a determination of the EB spot ovality may be made. Most electron guns do not generate a truly circular spot. Welding performance may be altered by the orientation of the major axis of the oval with respect to the welding direction.

Estimation of beam current

By integrating the area inside the energy density distribution of a given EB a measurement of the corresponding beam current may be made. This measurement may provide a useful double check for beam current calibration if there is any doubt regarding the accuracy of the machine's internal measurement systems.

Estimation of beam diameter

By analysing the energy density distribution of a given EB, a measurement of the beam diameter may be taken. Variations in this measurement will indicate some anomaly with the set-up of the machine.

Each EB welding machine will have its own characteristic beam fingerprint. This changes with variations in cathode condition and electron gun set-up. To determine this fingerprint, a series of energy density distributions of EB could be made in relation to systematic variations in the electron gun set-up.

In this way the effect of variations in electron gun set-up can be quantified in a relatively short period of time. If the probing system is used routinely before welds are made on production components then, over time, a correlation between the energy density distribution of the EB, the electron gun set-up and any variations in welding performance can be obtained.

This information may be used to establish the allowable variations in the energy density distribution of the EB for various types of components and weld.

In this context TWI is currently organising a Group Sponsored Project to investigate the relationship and correlation factors between gun set-up, probe traces and welding performance. Based on this analysis a code of best practice will be developed for monitoring gun condition and controlling welding performance.

The proposed GSP will enable EB welding machine users to implement novel quality assurance techniques, which potentially offer substantial cost reduction and defect prevention for production industries.

References