Cantilever versus Gantry Style Waterjets. Which one should you choose?

We want you to think of your WARDJet as "NOT JUST A WATERJET..." and we have already done the design, thinking, and planning to give you the foundation you need in a waterjet cutting system to do things that were simply never before possible with a cantilever style waterjet system.

Click to enlarge +
Typical Cantilever style waterjet	WARDJet Gantry Style Waterjet (Ma)

If you cut 2 parts at the same time, you just doubled your income revenue without doubling your cost.

You still have one machine, one operator, one program, one area of allocated floor space, one set of overheads. If you cut with 4 heads, you just quadrupled your income...(You will need the pump capacity to drive 4 heads). Then there are customers that are running several WARDJet waterjets with 15 heads on each system. For obvious reasons! But none of them have cantilever systems. Why? Let's have a look.

The waterjet industry has been inundated with cantilever style waterjet cutting systems in the last 10 years. They are low cost and certainly have forged a place in the industry. All appearances are that cantilever style waterjets far outsell gantry style machines.

So if it seems everyone is doing this, and you are looking to buy a waterjet, you must ask yourself, is this the way to go?

Yet one look at WARDJet waterjets shows an absence of cantilever style machines. So why has WARDJet not built a cantilever style system? There is surely a reason for this decision. A decision that flies in the face of what most manufacturers are doing. And that's what our challenge to you is today.

An interesting fact to consider is companies building cantilever systems often offer a gantry style as a "top of the line" system. So what is it that makes a gantry "top of the line"? What are you compromising when you go for a cantilever system? Read on and you will soon know more about this subject than most people who are in the industry!

Comparison of Cantilever versus Gantry Systems

The following "gantry" comments are based on a WARDJet gantry style waterjet. Please note there are many gantry style machines that are so lightly built that most of the statements below would not apply.

Why are Cantilever Waterjets Built?

Low Cost / Price

The waterjet industry is extremely price conscious. Everyone wants to spend as little as possible and achieve the end goal in an acceptable manner. A cantilever system is certainly substantially cheaper to build than a gantry. Building a cantilever system is typically justified by pointing to the low reactionary loads exerted by the waterjet cutting process. The typical up-thrust from a cutting head is less than 20 lbs. There are minimal horizontal forces and loads on the motion system. This allows for the construction of a lightly built motion system.

A cantilever system requires:

• Fabrication of two beams - the back beam and the cantilever beam. (Gantry needs 3: two side beams and one cross beam)
• Machining of 2 beams. (Gantry 3)
• With 2 beams, there are only two sets of everything needed for motion, so two sets of linear rails, drive system such as rack and pinion, ball screw, belt, linear motors, and bearings. (Gantry has 3 sets of everything)
• Only 2 motors, drives and cables are needed - one per axis. (Gantry 3)

So if you only have 2 of almost everything as opposed to 3, it stands to reason that approximately 1/3rd of the cost is removed from the construction of the machine.

But what, if anything, are you sacrificing, when you go to a cantilever system?

Lower Accuracy Needed in Machining of Beams

Please note, the implication here is not that all cantilever systems are built to lower standards. This chapter simply explains how it is not necessary to be paranoid about accuracy of machining of the beams on a cantilever system. The explanation follows.

Cutting With One Head Encouraged on Cantilever System

Generally, cantilever systems are sold with one cutting head. If more heads are desired, they have to be mounted on a spreader bar. A future CHALLENGE will focus on the difference between mounting cutting heads on spreader bars and independent Z carriages.

Now, when most new machines are built, inaccuracies in the build of the machine can be compensated to some degree through mapping of the system. So if the beam is not quite straight, this can be measured using lasers. The deviation from actual position is then entered into the controller so when the cutting head reaches a specific position, the electronic compensation combined with the actual position on the beam, will result in true positioning of the cutting head. This is a quite acceptable practice and we would encourage it as long as the controller is reliable. If the controller was to lose these compensation figures, your machine would immediately fall back on the actual mechanical tolerances of the build of the system.

So, the point here is that theoretically, as long as the positional method used in the build of the gantry is repeatable, one could take string and tie it to the beam, mount it on low cost tubing, and using laser compensation in the controller, the machine could be quoted to holding close tolerances per axis. And again, this is quite acceptable and the end result could be very good as long as you are only using one cutting head. Hence the reason most cantilever systems are sold with single heads.

However, as soon as soon as you go to more than one head, you run headlong into major obstacles.

Multiple Head Cutting Avoided on Cantilevers

If you want to cut with multiple cutting heads at any stage on a cantilever system, relying on the practice of mapping out inaccuracies in manufacturing falls apart. The reason is that mapping allows compensation for the inaccuracies of a single point on the X and Y axis. WARDJet systems have the ability to map a plane or volume of space in the cutting envelope using the FARO Laser Tracker, but very few waterjet manufacturers own such calibration equipment. We suggest you inquire as to what calibration and verification equipment is used for mapping of a system prior to a purchase. We also strongly suggest you look for plane or volumetric mapping capabilities. Typically a laser interferometer is used to map each axis individually for positional tolerance in a straight line per axis. The X and Y position is then an interpolation calculated in the controller as the tool path is generated for the cutting head.

What on earth does this mean you are probably asking! Well, as soon as you add a second head, it is not possible to compensate for the position accuracies of the second head. What you have to do is compensate for one head, and then rely on the mechanical accuracy of the machining and build of the structure to position the second head.

So let's say a cross beam has a bow of 0.005" over 12 feet, with the maximum deviation of the bow in the middle of the beam. That means the center of the cross beam is 0.005" behind the ends of the crossbeam. Now if you have one cutting head at the side of the machine, and the second cutting head 6 feet away right in the middle of the cross beam, as you start to cut a circle of 6 feet diameter with each head, you will have two different circles as each head moves along the cross beam. The first head will be moving back toward the 0.005" as it approaches the middle of the cross beam, while the second head will be moving forward as it approaches the end of the beam. The result will be two different shapes even though they were cut on the same machine at the same time.

The only way to address this is to ensure the accuracy of the machining and build of the structure. And this costs money. Mapping cannot compensate for this. There is no shortcut. Your build and machining practices need to meet the highest standards.

So when you put 2 heads on a spreader bar on a cantilever system, say 48" apart, if there was a bow in the cantilever beam of 0.005 in the middle, this could result in the spreader bar being twisted or rotated by this amount as it moves. If we calculate the effect this would have on a cutting head 24" away from the center of the travel of the beam, the cutting head would move off position. What is worse is the right hand side cutting head would go backwards and the left hand head would move forward magnifying this error.

This would not be true if the beam was machined to extremely high tolerances.

WARDJet's Machining and Build Practices

Everything you read about WARDJet encourages the use of multiple cutting heads on one cross beam -- heads that can be spaced close to each other and then moved over 12 foot apart. And we don't stop at multiple waterjet cutting heads. We encourage multiple processes, like drills, tapping, height sensors, crash sensors, 5 axis cutting heads, marking, cameras and other accessories that we have not released. But even more, we encourage our users to find other processes they would like to add, like liquid dispensing, light duty routers, bar code reading, light assembly and automation. We want WARDJet owners to see their waterjet as "NOT JUST A WATERJET..."

Click to enlarge +

Click to enlarge +

And this would be impossible with a cantilever design.

We simply lose all our flexibility and a large amount of our creativity when we limit ourselves to a cantilever system. By building heavy duty gantry style machines, we keep all doors open to whatever our imagination or customer opportunities present. We are free to think outside the box with all WARDJet waterjet users. We cannot think of building a waterjet any other way.

Fabrication Of WARDJet Gantry and Stress Relieving

First, all side beams, cross beams, risers, and, in some cases the tank, are heat stress relieved or vibration resonance stress relieved. The purpose of this is to relax any stresses induced during the welding of the different steel components. If you can envision welding the different parts of the side beam together, we get localized concentrations of heat around the weld. The steel will heat up and change its characteristics relative to other areas where the steel did not heat up as much. This results in stresses being induced into the fabricated beam. Over time, these stresses may reduce and as they do, the fabrication will change its shape. Let's say we machined the beam prior to stress relieving, and recieved excellent results in our machining. If we were to wait a year and then measure the finished product using a FARO Laser Tracker, we would notice that it was not the same, because the beam had not been stressed relieved.

So stress relieving of the parts and components is critical.

Machining of the Beams

The accuracy of the CNC mills used to machine the waterjet beams will determine to a large extent the mechanical accuracy of the waterjet. If the mill has a bow in one axis, this will be passed onto whatever work is machined by this mill. It might be worth inquiring as to the accuracy of the mills being used to machine the beams.

Other Waterjet assembly And Installation of Linear Rails

So you have your beams machined, but everything ultimately runs on a rail or guide of some kind. It is imperative the rails are absolutely straight, parallel and true. This is where you have to ask, how are the rails installed, torqued to the beams, and measured for straightness?

Generally machine builders will use a lip machined into the support surface of the rail to push the linear rail up against. The presumption is that the edge of the lip is straight and meets all the requirements from an accuracy point of view. The linear rails are then tightened down to the beam. Of interest, is the simple torque applied to each bolt going through the linear rail will induce a bow into the beam if each bolt is not shaped correctly to minimize this effect. Most manufacturers then rely on the straightness of this lip to determine straightness of travel.

WARDJet Assembly and Installation of Linear Rails

WARDJet again does this a lot different than most manufacturers. We do not machine a lip on the cross beam to position the linear rail. Then how on earth do we ensure our rails are straight? Well we go overboard I guess.

Once the beams have been stress relieved and sand blasted, they go to the mill for machining. The mills we use are all calibrated and certified with documentation to verify accuracy. As WARDJet owns certified FARO Laser Tracker equipment, we can of course routinely verify accuracy of all our equipment.

Linear Rails Installed on the Mill

The beam is then taken to powder coating and here is where WARDJet takes a radical turn from the normal. After powder coating, we put our beams back under the mill and install the linear rails onto the beam. We do not use a lip but rather set the rails and verify the position and straightness of the rails using the mill. As long as the mill is accurate, which we calibrate, we know exactly what the straightness of the rails is on each of our beams. Every rail installed on a cross beam is measured for accuracy before it gets close to installation.

Laser Confirmation of Linear Rail Position and Pinning of Rails

Once we are happy the rail is straight, we then will use a laser to verify this one more time. Once all requirements are met, the linear rail is pinned into position onto the beam. The method of pinning then allows us to be sure that the rail is exactly where we need it to be. We do not rely on pushing the rail into a machined lip but rather rely on calibrated and certified laser equipment. So yes, we go to great lengths to ensure accuracy and straightness of our linear rails.

And because of this, we then are comfortable encouraging those who trust WARDJet to look after their waterjet needs, to use their WARDJet as "NOT JUST A WATERJET..."

Squaring a Cantilever Verses a Gantry

When you look at a cantilever style machine you effectively have a "T" with the one side being along the back of the table and the cantilever beam projecting out over the working area. When any machine is installed it needs to be squared such that the X axis travel is exactly 90 degrees to the Y axis travel. When you have a cantilever, the way to get these axes to be 90 degrees to each other is to take measurements and then to make physical, mechanical adjustments to the machine until the beams are 90 degrees to each other. The system is then bolted down to lock this into position. So a physical, mechanical adjustment is required. If this goes out of whack at some stage the beams will have to be squared again in this method. Some manufacturers may assemble a machine and then provide pins to fix the relationship between the two beams. You may want to ask how this is done if you are considering a cantilever style machine.

There is no way to use a set of measurements taken, do a calculation, enter data into the controller and literally adjust for squareness in minutes.

Many controllers running gantry waterjets have no way to adjust for squareness and they too require some kind of trial and error adjustments with a new reading taken between each adjustment until a satisfactory reading is achieved. It is wrong to assume that a new waterjet has a method of squaring the system other than with a trial and error process. You must ask and check how the waterjet is to be squared upon installation.

WARDJet Gantries Squared With Ease

We consider the need to square a WARDJet gantry as critical. If the machine is not square everything else will fall apart. It will be impossible to produce good parts both from a squareness and circularity point of view.

WARDJet gantries are unique in the way they are built. Each side of the cross beam is machined to high tolerances and rests on an equally highly machined surface on the risers on each side of the machine.

The one side of the beam has a hole to mate with a machine pin. The other side of the beam has a slot that allows the pin in the top of the riser to slide inside the slot. The two surfaces are then held together with 4 bolts. What is unique is the mating surfaces of the riser and cross beam is greased so the two surfaces can move relative to each other. There are many reasons we do this which are proprietary to WARDJet and advantageous to our customers, but one we will share is how easy this makes it to square a machine.

We literally do the following:

• With the bolts tight, we perform a ballbar and electronically check squareness, flatness of the machine, mismatch of motors and circularity. This is done when we build the machine to ensure it meets our high standards, and then always performed on site.
• WARDJet software then performs calculations in seconds and generates code that will lock the right hand side motor in position.
• The bolts are loosened and the left hand side motor is then activated and the left hand side riser and side of the cross beam will move accordingly. This will move the cross beam into position so it is square.
• Another ballbar is done to check the machine is square and that all the required specifications are met by the gantry.
• Results are electronically generated. There is no fudging the numbers.
• WARDJet has a backup of the actual status of your machine and a copy is given to our customers as a back up

Advantages of WARDJet Squaring Process

Unfortunately, reality is that your machine is likely to go out of square at some stage. Causes of this could be reversing the fork lift into a side beam, hitting it with a crane or fork lift, or failure of the motors in some way - like missing encoder counts.

Whatever the cause, data generated with a simple test done on site by the customer, enables WARDJet to square the X-Series controller machines remotely. You literally could stand back with your hands folded and watch this and then verify the successful results. With older waterjets, we guide the customer on how to address this easily, typically without having to leave our facility.

Deflection of the End of The Cantilever Beam

When you were still a child, it did not take long to figure out that if you hang something heavy on the end of a stick or fishing rod, it bends. Sometimes a lot. Scooping up that spoonful of ice cream with the plastic spoon was interesting as you could watch the spoon bend and on occasions, snap. That' cantilever action you are familiar with. Your hand is the one end of the gantry and the fish or ice cream is the cutting head at the other end. If you are trying to hold accuracies of 0.001" you had better be sure to ask for a finite element analysis of the deflection of the end of the beam. Then ask for the same calculation to be repeated again as you add more cutting heads and load onto the system.

Vibration of the Cutting Head

It is possible, and we have seen this first hand, where the deflection of the end of the beam ultimately resulted in a vibration at the end of the cutting head as it moved up and down. This is not common, but it does happen with some systems so again, check the design.

Deflection of WARDJet Gantry

WARDJet has finite element analysis predicting the effect of additional loads and torque induced in the beam at multiple locations on the cross and side beams. We know what our results are and we are pleased with them.

By supporting the cross beam on both sides, due to the way our cross beam is built and designed, our deflection is minimal as we add 4, 8 or 12 cutting heads and drills. What we do with WARDJet waterjets was simply not ever before possible. Then, as we own FARO Laser Tracker equipment, we are able to verify this with actual measurements.

And we do this with you in mind. We want you to think of your WARDJet as "NOT JUST A WATERJET..." and we have done the design, thinking, planning and laid the foundation for you to let your imagination go and expand your horizons.

The Challenge

So if you want to prove to yourself what could be right for you and your application, walk up to a machine and do or request the following:

• Reach out and grab the cutting head and put your whole body weight into it and see if you can see any movement in the tip of the cutting nozzle or focusing tube. This will typically be more prevalent on a cantilever system. You will also find some lightly built gantries will move more than they should. A fair amount of movement could also come from lightly built and constructed Z carriages.
• Position 2 cutting heads as far apart as they can go on a spreader bar on a cantilever and then cut two squares and two circles, then measure and verify the sizes, keeping the orientation of the parts the same.
• Do the same on a gantry system where the Z carriages can be positioned independently of each other. Have the one head at the end of the cross beam and the other in the middle of the span and do the same test mentioned above.
• Always ask to see the laser interferometer results on a machine. Ask for the results of the accuracy of the beams prior to compensation and after compensation. This will show you what you have to start with as well as confirm that any inaccuracies have been taken out. (Make sure you have a back up of these values as they will be specific for your system.)
• Load the end of the cantilever by hanging on it and see how much it deflects.
• Bang the end of the cantilever with your fist and see if you notice any vibration on the system. An easy way to see this is to lower the cutting nozzle into the water. Then hit the end of the cantilever with your fist and look for tell tale rings / waves of movement in the water. Do the same with different cantilever systems.
• Ask for the torque loads the Z carriage can take when adding drilling and tapping heads. Then ask how many drills and tapping systems the cross beam can handle. WARDJet suggest a limit of 12, so 6 drills and 6 taps on one cross beam.
• Ask how many 5 axis cutting heads can be mounted onto the cross beam. WARDJet can install up to 9 Psy Winder 5 axis cutting heads on one cross beam.
• Ask for records of stress relieving of the various components of the machine.
• Ask for finite element analysis showing the results of loading the end of the cantilever beam.
• Ask for readouts of the machining results and accuracies of the linear rail placement on the beams.

We hope this gives you more insight and depth into some of the considerations that need to be taken into account when making the important decision on what style of waterjet cutting system you are going to trust to see you into the future.