Pressure Sensor for 3D printing

I am working on pressure sensor development. I am interested in applying our pressure sensor to monitor the pressure in the heated barrel of 3D printing, but i am not sure whether this is necessary for 3D printing. Is it is not necessary, any other sensors that the 3D printing may need?

Thanks

It could be real important to determine if plastic is being pushed out
and pushed out correctly.
Many times I have had air prints -- head moving around with NO plastic coming out.
Would be good to know filament is not exiting, stop print and recover.
Also to be able to determine if the flow is correct.

The sensor has to withstand the pressure and temperature found in extrusion printing

Looking foward to a inexpensive and accurate sensor.

Hi cozmicray,
Thanks for your reply. Do you know the pressure and temperature range? and where the pressure/tempreature sensor should be installed?

I suspect a vastly simpler sensor would accomplish about the same: paint a few lines on the idler bearing and use an optical encoder to check if the idler is moving as far as the hobbed bolt. Every time I've had a jam, the hobbed bolt grinds down the filament and the idler stops moving. This seems like a fairly easy way to determine if the extruder is working correctly.

The Temperature range 170 C PLA to 200 C ABS to 250 C Nylon and polycabonate.
The pressure sensor has to be in the hot end melt zone.
Pressure of solid filament probably not measurable.
But the push force of the solid filament may be of use.

1.74mm or 3mm filament to a 0.2 to 0.5 mm nozzle.

Enough pressure to force the viscoelastic molten plastic through the small nozzle at a high speed
that when the extruder motor is turned off, the plastic will keep flowing until that pressure is relieved.

You might want to look at Airtrippers project
Airtripper Extruder Filament Force Sensor
[airtripper.com]

I think this is measuring force of the filament pressing into hot end.

What pressures are in the extruder? Good question

So the pressure sensor must be able to survive in high temperature enviroment (170 C to 260 C) and detect very small force, is that right? Are there any commercial available such sensor in the market?

an absolute position encoder attached to the idler bearing would probably be a lot more useful than a pressure sensor in the barrel. the bearing would need to be grooved and have some sort of textured surface to make sure it doesnt slip on the filament.

aduy: With pressure or similar sensor, firmware can sense as force is gradually approaching hobbed bolt limits and possibly take counter-action (adjust temperature, slow down extrusion). With sensor on idler you get signal only when its too late.Though I think so, using tensometric sensor to measure axial force exerted on hot-end (already done, there should be topic somewhere here about it) should be much simpler and cheaper than constructing pressure sensor that can withstand hot-end environment.

Are there any advantage that if the pressure sensor which can withstand high temperature is constructed? should this sensor be very small to be installed in the barrel?

Rather than a pressure sensor, I've thought about replacing the extruder stepper motor with a servo motor. You could then use the motor current as a pretty good estimate for the extrusion force. Servos also have a lot of advantages in applications (like an extruder) where the required torque can vary considerably.

The force to push the filament is not the same as the pressure at the nozzle as a large proportion is taken up pushing the viscous plug at the transition zone.

A pressure sensor at the nozzle would be very useful because flow rate is directly proportional to pressure so you could retract to the point where the pressure was zero at the end of a filament run and then fast forward to the point where the pressure was correct for the desired flow rate at the start of the next run.

I have no idea how one would implement one at the hot end though, especially on such a small scale.

---

[www.hydraraptor.blogspot.com]

Hi nophead,
How small scale you think? i think the diameter of the barrel can be 3mm and the height of the barrel can be 4 mm, so the sensor needs to be placed inside a cone shape cylinder with a dimension of 3 mm and height of 4 mm, is that right?

monkeyA: You may use J-head as reference. There is more information on its wiki page. You may also review other hotend designs

Thanks Miso, i have reviewed the J head design, i see the cavity has a diameter of 6.5 mm and the nozzel head has a diameter of 2 mm, my question is: can the pressure sensor be placed inside the 6.5 mm diameter cavity or it has to be placed right on the head of the nozzel?

It's unlikely that you will be able to measure flow consistency from the melt chamber pressure because from the results I've produced from my own kit shows extruder filament force measurements are also influenced from outside the nozzle.

Perimeters use less extrusion force than in-fills (both set at same speed). In fact, any warping and filament clumping as well as different fill patterns will have some influence on melt chamber pressure.

Pressure balancing inside the melt chamber would be a great feature to have but it would probably work better in something like the filabot that extrudes into the air.

The best solution is probably to buy good quality filament. I've got some filament here that will leave strings no matter what the retraction distance I use, and the same filament requires the most force to extrude compared with other filament.

If I had to design a melt chamber pressure sensor, I would probably start with a vacuum probe.

why not calculate pressure based on feedstock diameter, then the sensor can be placed above the extruder in the cold end, and a delay by how many mm from extruder hot end allows pressure to be approximated, and even though it does not allow for automatic configuration of machine, it does allow it to work constantly even if feedstock varies in diameter. a sensor could be made out of conductive foam (electrostatic pads for electronic parts) above the hot end. changes in resistance can cause a change in flow rate after the distance is extruded between sensor and ectruder. it can calibrate from start up extrusion of 10-20mm of feedstock.

Been following this thread with interest, and I am coming at this from a position of being very new to 3D, but a good overall background in some of the areas that are involved, so the suggestion I will throw into the pot is that a possible way to get the information would be to find a way of mounting a strain gauge between the hot end and the extruder, or bowden tube end. This would sense the pressure being applied by the extruder to the hot end, which is not quite the same as the pressure at the nozzle, because of the transition zone changes, but it should be possible to derive the pressure with some fancy math, (which I don't have a hope of doing, so don't ask ) and this has the advantage of not compromising the hot end nozzle area, which seems to be the critical area of operation

I can see potential problems with a pressure sensor on the nozzle, one of which will be the potential for more stringing unless the retract filament distance is made longer, as the volume of molten filament will increase, it will have to be molten in order to allow pressure transmission. That also then increases the heat demand on the hot end, which may result in a slower response to the thermistor control, which could impact on the accuracy of temperature control of the filament. As I understand things, the transition zone for the filament has to be very carefully designed to prevent plugging, or jamming, so introducing a side exit from that area to connect to a sensor could cause as many problems as it solves. especially if the line to the pressure sensor has not liquidised when pressure is applied from the extruder. Maybe I'm reading more into the issues around the nozzle than are there, but from the reading I've done on the thread about my chosen hot end, I don't think I have, the instructions for setting up the nozzle and interface from the cold end to the hot end are very specific and detailed about ensuring that there are no gaps between the two, to avoid issues in that area.

The advantage of the strain gauge at the other end of the hot end is that it should not have any impact at all on the heating process, and the movement will be very small.

Hope I haven't muddied the waters on this.

---

Shore, if twas easy, we'd all be doin it

Irish Steve

Using a strain gauge to measure the extrusion force has already been done but it is very different from the nozzle pressure. At some speeds the force actually goes down as speed increases because the plug length reduces whereas nozzle pressure will always increase with increasing flow rate.

---

[www.hydraraptor.blogspot.com]

Pressure sensor will not work with heated plastic.

The plastic will get inside it, and cool, and it will not be heated properly again until the heat propagates through the plastic which will take a long time, rendering the sensor useless.

Sensors that would be of use 3D Printing.

- Incremental Encoder separate from extruder motor, to measure filament input.

- Electrical current sensor on a servo fed extruder. (This is your pressure)

- Electrical Dial indicator to measure on the fly filament diameter.

- High Accuracy range finder (optical probably better than acoustic) - Distance to print surface


I'm sure someone could design a idler/Sensor combo that could replace the existing idler on a Wade Extruder that could be both the Inc. Encoder and incorporate a measurement of it's movement for filament diameter.

Edited 1 time(s). Last edit at 09/03/2013 09:58AM by ShadowRam.

nophead Wrote:
-------------------------------------------------------
> Using a strain gauge to measure the extrusion force has already been done

Thanks, I missed that, it probably happend a good while ago, and I've only been following closely for a few weeks, as the plan to build a printer only came good recently.

Understand the problem, I guess I was thinking that getting close enough to the nozzle without upsetting the operation is going to be tricky, and even more so if the same nozzle is changed between filament types or colours, which could complicate things even more.

ShadowRam has also expressed slightly better than I did the worry I was having about "static" and less liquid plastic in the sensor path causing problems, unless the sensor is mounted very close to the nozzle on the heater, which then means issues with heat being "diverted" to it, rather than focussed on the nozzle,

This might be slightly off thread, but I will ask it here.

What sort of distance does it require for the filament to go from being solid to being extrudable, my thinking being that in order to use the tubular cylindrical heaters that seem to be the popular route right now, we are using a block of aluminium that the heater is mounted horizontally in, and the nozzle is threaded in to that block from one side, and the "interface" to the feed comes in the other side.

I was thinking that a possible alternative would be to use something that looks similar to a sewing machine bobbin in terms of shape, except it would be a solid centre, threaded to take the nozzle, feed input, and drilled to take the thermistor. The heating coil would then be wound around the outside of the hub, and could be whatever resistance is deemed appropriate in order to provide the heat required to acheive the demanded filament flow.

Going back to pressure sensing, depending on the thickness of the liquid layer, it might be possible to have a strain sensor concept on the nozzle and heating coil, with the interface being at the back of the transition zone, but even there, keeping the seal on the join, and getting accurate measurements of the pressure could be problematic.

If I'm going over old dead ground, shoot me down quick, I won't be offended

---

Shore, if twas easy, we'd all be doin it

Irish Steve

ShadowRam
How do you know a sensor will not work in molten plastic?
What kind of sensor? If you don't know how the sensor works
how can you say it will get plugged up.
I don't think this sensor will be a pitot tube.

"Dynisco's wide variety of robust 0-10 and 0-5 volt melt pressure transmitters; have been
designed specifically for harsh and rugged environments of the extrusion and polymer processing industries."
[www.dynisco.com]

Pressure Transducer for Plastic Extrusion
[www.hcs77.com]

I have a crewed drawing below, but it explains how you would sense diameter. the device is 1mm thick. the inner non conductive layer would be stretchable but slick and durable. Also another sensor is build into the bottom that senses feedstock movement. this is all theory, there are several physical limitations with material choices and reliability that need addressing.

about 20mm before the feedstock enters the hot end is where this is placed, the total volume going to the extruder is calculated in advance. the output flow rate is calculated base on the diameter of the feedstock on the fly.

also with a conductive bottom to the conductive foam, another sensor could measure some sort of downward force, and know if the feedstock is still going downward in direction. if it is then the feedstock has not slipped, or if it starts to slip the entire speed of printer can be reduced to allow feedstock to continue without fail.


so this simple idea of a sensor does 2 functions for reliability. it lets a machine know when feedstock starts to slip and each time slows down the machine by say 10% for a minute to reduce flow rate and back pressure and it monitors feedstock diameter to adjust flow rate.

so even though you don't know the actual pressure, it always keeps the machine within the pressure range that it can still extrude plastic




Edited 9 time(s). Last edit at 09/03/2013 05:19PM by jamesdanielv.

Why is filament diameter needed in the sensing of extrusion pressure?
My 1.75mm filaments vary from 1.70 to 1.87mm.
and I place the average diameter into my Slic3r program.
Would filament diameter be used for volumetric control of extruded plastic?

?????

it is a way to know when pressure is building up inside the extruder. it does not directly measure the pressure, just the effects that change pressure. if flow rate is too high and pressure builds up to where feedstock slips, then the issue is too much pressure. adjusting flow rate until this no longer happens would be a way to control pressure inside nozzle.

So you believe that the filament diameter, above the hotend meltzone, increases with pressure?
I guess I'll have to press real hard axially on some filament and see it change diameter?
Perhaps that why extruders jam a lot, they don't account for ballooning of filament.
Your sensor perhaps will have to run with change in diameter in angstroms?

how about using a current sensor to monitor the motor current being used by the pololu board itself. i dont know about the coding needed in marlin but the wiring should be simple enough.

using a board like this: [www.aliexpress.com]

i dont know about sensitivity though, 185mV/A of sensitivity, what does that means? i suppose you divide the mV/A by the voltage you are using, that would mean 15.4 mA of sensitivity at 12V.

Here's what to do.
Those who have any ideas as to add marlin?

[basdebruijn.com]



Edited 1 time(s). Last edit at 05/13/2014 01:17AM by radus.

probably a calculation of distance before a travel move. if a distance is within a certain amount based on buffered moves then extruder is turned off in advance.

but i don't see how this works better than a controlled retraction.

i think it would be more benefit to have software that would move that extrusion and layer change to an inside layer, so it does not effect visibly the manifold solids outside layer.

Edited 1 time(s). Last edit at 05/13/2014 05:09AM by jamesdanielv.

i dont know if this is close to what they are doing with linuxcnc, but i found this:

Modified Slic3r for Bowden

i'm not sure about my settings, but it does seems to help a lot.

Pressure sensor DYNISCO for molten plastiscs extruder.
working temperature 400 degress.
[www.dynisco.ru]