Twin-Screw Extruder Gearbox Retrofit: Check Shaft-Center Distance First

A twin-screw extruder gearbox replacement usually starts as a torque and ratio exercise. The existing unit has a known output speed and a known power rating, the process needs a certain specific torque, and the search begins for something that meets both numbers.

Then the drawings arrive and the project stops, because the two screw shafts on the extruder are a fixed distance apart, and that distance is not negotiable.

This is not a manufacturing tolerance problem or a matter of picking the right option from a catalog. Shaft center distance on a twin-screw machine is a geometric property of the barrel and the screw set, fixed before any gearbox exists. A replacement unit either has output shafts in the right places or it does not fit, regardless of what it can transmit.

Why the Center Distance Is Fixed

In the co-rotating intermeshing arrangement, which is the case this article works through, the two screws sit parallel in the barrel, turning in the same direction with their flights passing through each other. That requires a specific relationship between screw diameter, flight depth, and the distance between shaft axes. Change the center distance and the screws either collide or stop intermeshing.

The arrangement has to be confirmed before any of this is applied, because it is not the only one. Counter-rotating machines exist, where the screws turn toward each other or away from each other and the drive has to impose that relationship. Non-intermeshing, or tangential, machines exist, where the flights never meet and the shafts are free of each other geometrically. Conical twin-screw machines exist, where the shaft axes converge and the center distance is not even constant along the barrel. Each puts different constraints on the gearbox, and the drawings for the machine in hand are what establish which one is being dealt with.

Published patent literature on twin-screw machines describes this as a fixed center distance between the parallel screws, treated as a given in the machine's mechanical definition. It is a barrel and screw set parameter, not a drive parameter.

The consequence for the gearbox is stated plainly in the engineering literature on extruder transmissions. Because the screw center distance constrains the output shaft spacing, the gearbox center distance is likewise constrained, and has to be made small. That is the source of the whole design difficulty in high-torque twin-screw gearboxes. It is not that the torque is unusually large. It is that the torque has to be transmitted through two shafts whose separation is set by something else entirely.

So a gearbox for a twin-screw extruder is a device that must satisfy two external constraints at once: a torque and speed requirement from the process, and a shaft spacing requirement from the machine geometry. The first is negotiable in the sense that you can choose a different torque class. The second is not.

What the Architecture Does to the Torque

Once the center distance is fixed and small, the gearbox architecture is forced into a compromise, and understanding that compromise makes the compatibility question much clearer.

One arrangement seen in this class is the parallel three-shaft layout, where power enters on the input shaft and is divided between two output shafts. That is one architecture among several, and how the division is made and how evenly is a property of the specific design. It has to be read from its drawings rather than assumed: a unit built for an even division and one built to carry an uneven load are different internal arrangements, and the total torque figure does not distinguish them.

Closely spaced output shafts can constrain gear dimensions and capacity, but the limiting element cannot be inferred from shaft spacing alone. Depending on the actual architecture, Materialien, Geometrie, Lager, Schächte, and rated conditions, the governing limit may differ. Establish it from the design and manufacturer-approved ratings for the specific gearbox.

This produces a situation that surprises people comparing datasheets. Two gearboxes with the same nominal total torque may have quite different internal architectures, different split ratios, different limiting components, and different behavior when one screw loads up more than the other. The number on the nameplate is a summary, and the summary hides the structure.

Some manufacturers publish a specific-torque or torque-density figure. Compare such values only when each manufacturer states the definition, Einheiten, and basis used. Without aligned definitions and rated conditions, the figures do not support a cross-manufacturer ranking or a universal industry tier.

The point for a replacement project is that torque class and architecture are coupled rather than independent. The internal layout sets how much torque can be passed through the output shafts at a given spacing, so a requirement for more torque cannot usually be met by keeping the same layout and changing nothing else. That coupling is the reason the geometric question has to be answered first.

The way into that question is the screw drawing, not the old gearbox nameplate. What the screw set needs follows from screw diameter, screw geometry and the process requirement, and those come from the barrel and screw documentation. Working backward from the rating of the unit being replaced is working backward from whatever compromise was made when that unit was chosen, which may have been a space decision or a delivery decision rather than a duty decision. The nameplate records what the machine has, not what the machine needs.

Not Every Twin-Screw Drive Is 1:1

An assumption that causes trouble in replacement projects is that the two output shafts always turn at the same speed. Published patent literature describes twin-screw machines in which the two screws run at different speeds, a configuration adopted to reduce degradation of certain materials.

That matters for two reasons. Erster, a gearbox that produces two different output speeds has a different internal arrangement from one that produces equal speeds, and the two are not interchangeable. Zweitens, if the existing machine runs the screws at different speeds, a replacement gearbox with a 1:1 output will change the process, not just the hardware.

The rotational direction relationship is a separate variable again, and it is tied to whether the screws are intermeshing or non-intermeshing, and to the intended conveying behavior.

The Comparison That Actually Decides It

Most of the rows in this table are the questions asked of a high-torque gearbox for twin-screw extruders at the enquiry stage, and the ones that cannot be answered are usually the ones the enquiry has not yet included.

The productive way to approach this is a drawing comparison, done before any discussion of torque. The question is not which gearbox is better. It is whether the existing installation and a candidate replacement share the same geometric frame.

Rows one and twelve are the pair that most often decides a project late and expensively. Center distance measured cold on a gearbox that runs hot is a number that describes a different machine from the one in service. The barrel and the gearbox expand by different amounts, and the mounting arrangement determines how that expansion is accommodated. A unit whose cold center distance matches but whose thermal growth differs from the original can develop a growing misalignment at the output coupling as it comes up to temperature, depending on how the two are connected and on how expansion is compensated between them. It is a row to check, not a prediction that the fit will fail.

Where Measurement Goes Wrong

The measurement itself is worth a note, because it is done by more people than are equipped to do it accurately.

Center distance is the distance between the two output shaft axes. On a unit in service, that means measuring between the two shaft surfaces, correcting for the two radii, and doing it at more than one axial position to detect any parallelism error in the existing installation. If the existing gearbox is being replaced because of a failure, the housing may be distorted, and measuring it will produce the distortion rather than the design intent.

The reliable reference is the extruder, not the old gearbox: the barrel's screw bore center distance, taken from the barrel drawing or measured at the barrel end face. That is the number the new gearbox has to match. The old gearbox is evidence of what was there, and may itself have been a compromise.

Where the barrel drawing is missing, the measurement has to be made on the barrel itself, and it has to be made properly. With the screws withdrawn, indicate each bore in the end face and derive the two centers, or use expanding plugs in both bores and measure between them. Take the reading at more than one axial position along the bore, because a pair of bores can be parallel at the face and out of parallel further in, and the gearbox will inherit whichever condition the barrel has. Record the barrel temperature at the time of measurement; a bore measured on a cold barrel is the cold figure, and the unit that bolts to it will reach a different temperature in service.

Shaft end form is the second frequent miss. Two units can have the same center distance and different output shaft splines, different shaft end diameters, or different axial positions of the shaft shoulder. All of these are visible on a drawing and none of them are visible on a datasheet.

What the Screw Set Wants From the Drive

One more thing belongs in the comparison because it constrains the answer: the gearbox is not free to impose its own idea of load sharing on the screws.

In an intermeshing machine the screws are mechanically coupled through the melt and through their own geometry, and the torque each carries is a function of the process, not of the gearbox. In a non-intermeshing or tangential arrangement the two shafts are more independent and can run at different speeds, which means the instantaneous loads on the two output shafts can differ substantially.

For a gearbox, this is a question about how the two output shafts are connected internally and what happens when one of them sees a higher load than the other. A design with an even split assumed is a design that will behave differently from one built to accommodate an uneven one. This is not something that can be resolved by looking at the total torque figure.

The Order the Two Questions Have to Be Asked In

The geometric question comes first, and it is answered from the barrel drawing or, where that drawing is missing, from a proper measurement of the barrel bores to the accuracy the fit depends on. That measurement is a metrology job, not a check with a tape measure: the tolerance that matters is the tolerance on the fit, and it is measured with the instruments and the method the machine's own documentation calls for. A unit that meets the torque requirement and misses the center distance is not a candidate at any price, and finding that out before the order costs nothing except the time to measure properly.

The torque question comes second, and it is the harder one, because it depends on what the screw set is being asked to do rather than on what the existing gearbox says. A unit that fits perfectly and is in the wrong architectural class for the duty is only discovered after installation, running at process conditions it was never designed for, and that failure costs a season of production.

Both failures come from the same root, which is treating a twin-screw gearbox as a torque-rated commodity rather than as the mechanical interface between a motor and a fixed screw geometry. The first filter costs nothing and it eliminates most of the field. The second one requires the process data, and no datasheet answers it.

For a retrofit enquiry, the four documents that make the geometric filter answerable are the existing gearbox drawing, the barrel drawing with the screw bore centers, the screw shank drawing, and the motor and coupling arrangement. Where the barrel drawing does not exist, the bore measurement described above stands in for it. Those can be sent through the contact page, and the fit question can be settled before the torque question is opened.