Below you can find a Q/A list based
on input from the sales department.
Please feel free to contact us with any additional
questions, and in case it is of general interest, it will be added to this
list!
Q1 :
Compare SLAB/RF to DC excitation.
Which is best for what
application ?
Q2 : Gas consumption SLAB is much lower
compared to DC, only one bottle of gas per 12 months.
Q4 : A small beam diameter (SLAB) is better
than a large beam diameter.
Q6 : The wall-plug efficiency of SLAB and RF is better.
Q7 : RF and SLAB technology occupy less
space than a DC laser.
Q8 : Mirrors contaminate more rapidly in
the DC case.
Q9 : Maintenance on a SLAB laser is cheaper
and simpler.
Q10: RF lasers are more compact and thus easier to build into a machine.
Q11:The SLAB laser mode pattern allows much faster cutting at the
same power level.
Q13: An RF laser can be pulsed at high frequency and is faster to change power level.
Q16: International service available ?
Modern DC high-voltage supplies as
the one developed for the HLT4000 are all-solid-state high-power transistor
(MOSFET) switched-mode power supplies which makes them both conceptually
simple, reliable and very efficient electrical-to-electrical (>95%). They require minimal cooling (only the
transistor mounting plate is water cooled) and hence decrease the demand on the
chiller unit.
RF power supplies are basically very-high-power radio transmitters that
require careful matching of the generator and transmission lines to the load
(laser discharge) and operate typically at 65% efficiency, which means that
they can’t be neglected when considering the capacity of the necessary chiller unit (they basically need a cooling unit of their
own. As an extra remark it should be
noted that RF also necessitates using deionised cooling water which further
increases the demands on the chiller unit and tubing
(all-stainless steel construction) and hence the price. A DC laser can be cooled using regular
demineralised water with some standard additives. However, for large-area ‘thin’ discharges
like in a SLAB laser RF is the appropriate (only) technique for obtaining a
stable glow discharge, hence its use.
The use of RF for cylindrical discharges as in fast-flow lasers has
grown historically from the concern of contamination of the mirrors by
sputtering products originating from the electrodes (for DC the electrodes are
in contact with the laser gas, for RF the power is fed capacitively
through the tube walls, which poses other problems such as deterioration of the
glass at impurity locations and eventually leakage through the tube
walls). For modern day DC excited lasers
this problem has been overcome by the design of the electrodes (shape,
material) and the gas flow circuit (mirrors in no-flow zones). To summarize, the typical arguments heard in
favour of RF over DC were true up until the end of the eighties, beginning of
the nineties, but are obsolete to-day, especially considering the completely
novel bottom-up design of the HLT4000.
Expressed
in m³ of gas this is true. However,
expressed in €, the picture becomes drastically different. In order to stabilize the large-area, thin
sheet SLAB discharge, it is necessary to add the very expensive noble gas Xe to the laser mixture.
An actual user of a DC025 (2.5kW SLAB) reported a gas price of
€1000,-/5000h (single bottle exchange).
Now, typical prices for the three pure gases He, N2 and CO2
are €12.38,-/m³ , 1.2,- and 3.17,- respectively (high-quality LASAL range
of gases by Air Liquide, Belgian prices). A 77/20/3 mixture of these at a continuous
replacement rate of 20Nl/h (1000Nl=1m³), results in a cost of
€987,-/5000h. Give or take a few € this
is identical to the (lower power) SLAB laser.
In fact, due to the special alloy from which the high-voltage electrodes
are made, they also act as catalysts to regenerate the mixture. So, the refreshment rate is no longer used to
counteract dissociation but rather to compensate for the inevitable small
remaining leakage of air into the system.
In laboratory conditions the HLT4000 has run sealed off for prolonged
times (several hours) with negligible power loss. So, allowing a small, continuous refreshment
rate will ensure operation to specification even in the presence of small leaks
that would cause unnecessary downtime in case the laser would be run
sealed-off. We believe this is a
sensible, industrially robust compromise.
Finally, it should also be noted that the SLAB laser mixture contains
non-negligible amounts (percentages) of
extremely poisonous carbon monoxide, which may require special
precautions in case the laser is used in an enclosed environment.
This may hold true to a certain
extent for older, U-fold type of resonators (regardless of whether they are RF
or DC excited). They are asymmetrically
loaded by atmospheric pressure forces and tend to bend and/or twist (torsion)
when evacuated. Moreover, they are
connected to a base frame using several types of bearing combinations that try
to isolate the deformation of this base frame (as a function of time and temperature)
from the resonator structure. The
HLT4000 resonator however is a perfectly symmetrical square frame of invar bars
so the vacuum forces cancel out and thermal stability is guaranteed. Next, it is held in position by ‘hanging’ it
in front of the base frame, without the corners of the square (mirror
positions) being attached to the base frame in any way. Furthermore, all mechanical components such
as electrode mounts, elements of the gas flow circuit etc. are firmly attached
to the frame but flexibly to the resonator.
This way the base frame and the flow circuit are effectively decoupled
from the resonator structure which makes it mechanically very stable. Typically, after transport &
installation, no re-adjustment of the resonator is necessary. Concerning the issue of DC excitation being
out of date, it is only recently that RF supplies have gone ‘all-solid-state’
in contrast to the HLT4000 which was transistorized form the very beginning.
Actually, from an optical point of
view, it’s quite the other way around: the larger the beam the better (which is
why laser beams are typically expanded before being focused). A larger beam has a lower intensity on the
lens and can be focused to a smaller spot.
The small beam diameter typically associated with a SLAB laser has been
known to lead to problems of optics burning in etc. (which is also the reason
why they need to use a diamond window on the laser in stead of ZnSe, the intensity of the beam in the long axis is simply
too high (SLABs typically output a line-shaped beam
which is corrected using a spatial filter and cylindrical optics outside of the
resonator)).
True, but the high-frequency noise
associated with the high-speed gas flow through the fast-flow system is easily
isolated by putting standard sound isolating foam on the inside of the laser
cabinet covers. In normal operation
(closed cabinet) the HLT4000 is silent.
Not true:
Comparing RF excited fast-flow
lasers to DC excited fast-flow lasers two remarks can be made:
- RF power supplies are notably less
efficient than DC ones (they come with a separate cooler)
- Glow discharges are characterized
by a ‘positive column’, where discharge conditions are favourable for laser
amplification, and ‘boundary layers’ where the processes that keep the
discharge going take place. DC glow
discharges have only a small cathode voltage drop (some 200 – 400 VDC compared
to 20kVDC operating voltage) limited to a small area directly at the cathode
surface. In contrast, RF discharges have
a ‘sheath region’ as large as the surface of the electrodes ‘wrapped around’
the discharge tube, where a non-negligible amount of energy is wasted and
discharge conditions are unsuitable for laser amplification.
Comparing SLAB laser technology to
fast-flow technology a remarkable fact arises. For a
fast-flow laser, roughly speaking, half the energy consumption stems from the
blower, and the other half from the power supply (e.g. 20kW blower motor, 20kW
power supply, 4.5kW laser = 22.5%
electro-optic efficiency, 11% wall plug efficiency). So, normally speaking, if a SLAB design would
be as efficient electro-optically as a fast-flow design, and since it doesn’t
need the blower, wall plug efficiency of a SLAB laser should be double! But, from manufacturer spec sheets, it
can easily be derived that the wall plug efficiency of these lasers is worse
(e.g. quoted consumption of 54kW at 3.5kW output power)! The only conclusion is that the electro-optic
efficiency (a combination of discharge efficiency (how much power goes into the
upper laser level) and resonator extraction efficiency (how well does the
intra-cavity beam match the active medium)) is half that of a fast-flow design. Whether there is really less power coming out
of the laser for a given input power, or whether most of this power is cut away
by the spatial filter in the process of converting the highly astigmatic output
beam into something ‘gaussian-like’ is not clear. Most probably it’s a combination of
both. One argument that does hold true
is that a SLAB laser uses minimal power when idle (no beam) while a fast-flow
design has the blower running continuously.
This depends on what you call a
‘laser’. RF and DC fast-flow designs are
similar in size, if you take into account all components (e.g. the RF generator
may be physically separate from the resonator ‘box’, but still it is part of
the laser). Manufacturer data on e.g. a
DC035 3.5kW SLAB laser gives 2100x850x853mm for the laser head and
800x600x1900mm for the ‘control unit’ (=power supply and other
electronics). This totals to
2.43m³. The HLT4000 is a stand-alone
box, all inclusive (even the ‘external optics’ including anti-reflection
isolation, circular polarization, expansion & collimation), designed for
easy maintenance so all components are easily accessible, and still only has a
1750x1820mm footprint and a height of 1600mm (total of 5m³). It is true that size wasn’t a real issue when
developing this laser because originally it was developed for use in flat-bed
laser cutting machines where the size of the laser is minimal compared to the
cutting and shuttle table, fume extraction, chiller
unit, assist gas containers etc. Also,
from our year-long experience using Triagon lasers,
which were developed for minimum size, we decided to abandon this requirement
and opt for maximum accessibility of every single component to facilitate
maintenance (e.g. most components slide out on rails, every component is
accessible separately, that is without needing to disassemble anything
else,…). Of course this goes at the
expense of some space, but relative to the size of the complete machine this is
negligible. Moreover, the laser cabinet
was designed such that the same size will hold the 6 and 8kW versions. An OEM version of the laser (without the
‘box’) measures only 1600³mm³, or 4m³.
Mirrors get ‘contaminated’ by three
separate causes:
- condensation of ‘fumes’ (oil-like,
from grease or silicones or vacuum oil,…)
- interaction with the discharge
- contamination with material
sputtered from the electrodes
Concerning the first. This holds true for all lasers (maybe slightly
less for the SLAB design due to lack of blower) and is a simple matter of ‘good
practice’ when designing the vacuum circuit.
The HLT4000 uses a blower with high-vacuum compatible grease-lubricated
bearings, and the bearing housings are always kept at an underpressure
relative to the resonator avoiding any evaporated grease to enter the
resonator. Next, an oil-free membrane
vacuum pump is used. Finally, care has
been taken that the materials that come into contact with the vacuum are
chemically inert: anodised aluminium, stainless steel or high-purity ceramic
(aluminium oxide 99.7%). All O-rings are
VITON, even where it isn’t really necessary from a point of view of
temperature. No plastics, silicones and
such are used throughout the laser.
Interaction with the discharge is
probably less important in DC lasers because the power doesn’t radiate and as
such isn’t inductively or capacitively coupled to
nearby surfaces. Also, the mirrors in
the HLT4000 are on the grounded side of the discharge, and special care has
been taken to ensure proper grounding of the surroundings of the mirror
surface, so that no ‘excess charge’ can creep from the ground electrodes
towards the mirrors.
Contamination of the mirrors by
material sputtered from the electrodes has been known to be a problem in older
designs of DC lasers. However,
sputtering is a threshold process, so it can be minimized by careful design of
the electrode. The HLT4000 electrodes
are designed in such a way that increasing power doesn’t lead to an increase of
current density on the electrode, but rather to the discharge using a larger
part of the electrode surface, thus maintaining a constant, low current
density. Next, a highly
sputter-resistant alloy is used for the electrode which further reduces
sputtering. Finally, the mirrors are
located in no-flow zones so any small amount of material that would be released
from the electrode can’t be transported towards the mirrors.
Since ELAS has no ‘hands-on’
experience with SLAB lasers it is difficult to argue this. However, in view of the complicated beam
shaping optical setup that is needed to transform the raw, rectangular output
of the laser into a Gaussian-like shaped beam, it seems difficult to accept
that this procedure would be either simple or cheap, especially if these
special optics need replacement. To the
best of our knowledge this kind of maintenance is only possible in the factory
in
As argued above (Q/A7), there is no
size difference between an RF and a DC excited fast-flow laser. A SLAB laser is smaller by about half, but
this is negligible in view of the size of the machines where high-power CO2
lasers are typically used with. Also,
nearly all machines have the laser standing next to them, not ‘in’ them, which
further reduces the importance of the size issue.
Let’s analyze the situation. Firstly, there is oxygen assisted
cutting. In this case, cutting speed is
determined by the chemistry of the process (oxidation) and is the same for any
laser (that is, the maximum attainable speed is the same, so any laser can only
attempt to reach this same maximum).
Next, there is nitrogen assisted (‘fusion’) cutting. This case is further divided into ‘thin’
(<3mm thickness) and ‘thick’ (>3mm thickness). For larger thicknesses, a somewhat broader kerf is needed to be able to blow out the molten
material. Furthermore, a large depth of
focus is needed to maintain the intensity of the beam at a high enough level
throughout the material thickness. So,
sharp foci are useless for cutting thick material. For nitrogen-assisted cutting of thin
stainless steel an increase of productivity has been reported for the SLAB
laser, however, this would be only a small portion of the range of materials
and thicknesses to be cut by a typical machine set-up.
A modern (transistorized) DC power
supply can be pulse-width modulated just as an RF one and hence turned down to
very low power. Even if you need true DC
(CW) operation, the HLT4000 can be run at extremely low power due to its
proprietary electronic feedback stabilization.
In lab conditions, the laser has been turned down to as little as 50W
CW! However, this is just a curiosity to
demonstrate the stability of the power supply, glow discharge and resonator
structure, there’s no point in using a 4kW laser at 50W…but 10% (e.g. for part
marking) is quite possible.
As above (Q/A12). To further comment on this: an RF laser is
pulsed by necessity, it is the only way power can be regulated (basically it
runs at maximum power always, and the duty cycle is used to control
power). For low-power (=low duty cycle),
high speed applications this actually proves to be a problem because the ON/OFF
cycles will be visible in the result (be it cutting, welding or marking). Lastly, it must be taken into account that
the real limit on pulsing the laser beam is the molecular kinetics going on in
the discharge, and this is the same both for RF and DC. In short, it can be
shown that the discharge functions as a low-pass filter with a cut-off
frequency of about 3kHz relative to the input power modulation. This means that any pulsing above this
frequency is highly damped and will not be visible in the output power.
As a standard the laser communicates
via ProfiBUS with a SIEMENS 840D equipped
machine. In all other cases communication
is via purely digital IO, even for power modulation (built-in freely
programmable pulse generator with standard pulse library provided, so no analog setpoint needed for the
power regulation).
From our experience with other laser
sources that we interfaced to our laser cutting machines we learned that the
main problem facing the machine constructor is the adaptation of the laser to
the machine. This interface can range
from a simple phase shifter to an elaborate scheme that incorporates
back-reflection isolation, expansion and collimation of the beam,…However, to
perform this task the machine constructor typically doesn’t have the necessary
know-how in-house and must rely on expensive, external consultants. In case of problems this leads to a situation
where no-one takes responsibility and everyone blames the other party. Therefore, at the very beginning of the development
of the HLT4000 it was decided to integrate all necessary external, “special,”
optics with the laser source. The major
advantage of this approach is that all these special optics, which are either
curved or have special coatings so their correct operation relies strongly on
the exact alignment of the optic relative to the laser beam, have very well
defined positions in space relative to the resonator. The latter would be a difficult task if they
have to be positioned in space independently from the laser, and even if you
succeed in this, it remains an open question how stable this alignment will be
in time (vibrations, temperature changes,…).
On the resonator frame a reference surface is milled that is used as a
base for any combination of external optics necessary for the process at hand. ELAS has
dedicated software that will ‘fit the beam to the process’ and based on
that our mechanical designers can adapt the exact layout of the external optics
to the customer’s needs. The standard
configuration consists of a phase shifter to depolarize the beam and a
dedicated telescope that will expand and collimate the beam over a prescribed
range. For most applications this
configuration will be more than sufficient, the only ‘machine’ mirrors needed
to guide the beam to the process are standard, flat, zero-phase shift
mirrors. For processes that tend to
generate dangerous back-reflections, an extended version that will protect the
laser source by absorbing any back-reflected radiation before it can enter the
resonator is available. To summarize,
the HLT4000 comes with a ‘ready-to-use’ beam, but in the unlikely event that
the standard configuration will not suffice, ELAS has the know-how to adapt the
beam to the process based on customer requirements.
ELAS
is a pure production and R&D site.
Installation and servicing of lasers is done by