How to Manufacture Rolled Water Filter Cartridges?

Can you send pictures?

My first inclination would be to hire a mechanical design firm to produce the design, after becoming very familiar with your current processes and your future expectations for the new machine.

I suggest a mandrill with a hollow shaft. A smaller diameter inner shaft with a bulbous end slides inside and can be locked to two positions making the mandrill either 'long' or 'short'. Both outer and inner shafts rotate together. Fit a foam sleeve over the narrow part of the inner shaft. In the 'long' position the rubber is not compressed and it's O/D equals the outer shaft, say 0.9inches. In the 'short' position the rubber is compressed longitudinally between the bulbous end and the end of the outer shaft. This will cause it's diameter to expand and grip the core along the whole of it's inside diameter. The amount of shortening will control the gripping force. The outer mandrill should have a stop to bottom the core so that it is always precisely positioned. A mechanism similar to an industrial version of the latch found in a retracting ball point pen would suffice to shorten the mandrill. The operator pushes the end of the mandrill to clamp, and pushes again to unclamp.

Place the sheet in a shallow 10" wide x 45" long open ended tray to guide it as it wraps. Slide it forward to a position under a glue head and apply a bead of quick setting glue across the full width close to the leading edge. Slide it forward again until the leading edge is directly below the centre of the mandrill. Lift a spring loaded tension wheel into place to trap the sheet between the core on the mandrill and the spring loaded wheel. Rotate the mandrill 5 revolutions. The spring allows the tension wheel to back off as the diameter of the wrapped core increases. This will leave about 5" unwrapped. Run a second glue bead across the trailing edge of the sheet. Rotate the mandrill again to wrap the trailing edge onto the core. Fix shoulders on the tension wheel. They guide the sheet as it wraps but their edges are sharp enough to trim the sleeve to length. Now back off the tension wheel and slip the tube of webbing over the assembly and stretch to put it tight. Cut from the tubing roll with scissors. Trim the nylon to length by running the mandrill while holding the sharpened tension wheel shoulders up to the nylon.

When trimmed release the tension wheel, stop the mandrill, unclamp the core, slide off, dip the ends in glue, fix the caps. Don't forget to remove the trimmings from the mandrill before loading the next core.

The degree of automation depends on your quantities and how much you want to spend. You can semi-automate so the operator does most of the work. You would probably get 1.5-3 cores per minute from an experienced operator. Or you could stack cores in a chute. Slide them on with a pneumatic cylinder. Sense when they home against the shoulder and automatically apply the clamp. Feed the sheet with a drive wheel from a stack, similar to a printer paper drive. Apply the glue beads with an automatic dispenser. Apply the tension wheel with a pneumatic cylinder. Put the nylon on with expanding fingers, trim with a separate pairing blade, suck the trimmings away with a vacuum and run the lot with a PLC. Typical speeds attainable would be 10-15 cores per minute.

Thanks for the detailed scenario. The compressed rubber should exert a much better friction angle than our first attempt which used O-rings expanding on mandrill with compressed air. Sounds good.