Hotend theory

if you make a mark on your filament at the top of the extruder, then reverse it out, you should be able to make a farly good estimate at where the glass transition point is right?

you might have to allow for a little bit of stretching, but it should give you a fairly good idea.

we need to put all this stuff on the wiki with good diagrams and calculations.

Yes it normally parts where it is molten and then there is a section of slightly larger diameter that is where it has expanded to fill the bore but is not molten.

---

[www.hydraraptor.blogspot.com]

Thanks for the detailed reply nophead there is a lot there for me to think about.

I had been under the impression that PTFE was fairly vital in getting the smoothness of movement and low friction we need, but if not then it opens up the possibilities.

It also seems desirable to have as few junctions as possible as these are a point of potential weakness and leakage.

I think I am perhaps over thinking it due to the problems I had with the flawed extruder design I was sent with the kit, having not had any successful prints yet I really just want to make progress now. Until my new bar of PTFE arrives I think I will just bodge this one back together (modifying to eliminate the flaw). Do you think that I can get away with nylon 6 for the top 15mm or so of the barrel (mostly inside the extruder head) as that is all I have available to repair it right now?

George

nophead Wrote:
-------------------------------------------------------
> There will be a plug formed in the PEEK section as
> halfway though it it will be at about half the
> extrusion temperature, which is above the melt
> point. To get away with just PEEK is has to be a
> short section (and so requires a heat sink) and
> preferable tapered. See this design of mine:
> [hydraraptor.blogspot.com]
> xtruder.html
>
> I can't recall a hot end without a heatsink that
> doesn't use PTFE. You don't get a short transition
> zone without a heatsink, so it needs to be
> slippery.

I see. The V9 extruder has a pretty big aluminum base plate to which the upper end bolts into and a small fan, so that does the heatsink's job. This still sounds like a better idea than using PTFE at all after reading about the durability issues of PTFE at these temperatures and pressures.

George,
Yes, I think nylon would be OK for the top section.


ttsalo,
PTFE works if it is totally enclosed inside PEEK like the Makergear hot and the J-head, and it dispenses with the need for a heatskink and a fan making the extruder lighter. This is what I am trialling at the moment: -

---

[www.hydraraptor.blogspot.com]

Thanks nophead.

I have made the new parts now and went with aluminium for the top section in the end to act as a slight heatsink.

I also moved the heater block down to the very tip to try to help create the optimum temp gradient.

Need to silicone the heater resistor and thermistor back into the block and wait for it to set then will give it a go...

If it doesn't work well then I have plenty of aluminium and brass and stainless already, and some PTFE and tiny tiny (scary scary) drills on the way...


George

I am a noob and am building my first Prusa v2 (3" bigger). I decided to try out my ideas for a hotend. I have failed with the first TWO designs, so I am trying to figure out what went wrong with my hotend.

The first version's design issue was that it was massive. I did not even try heating it up as I used a bolt coupler for Aluminum block. I fixed the issue by switching to copper CPU heatsink.
The second version's problem was that the brass screw was 2.5" long. It was melting too much plastic. Also, I am using a 1Ohm aluminum encased 5Watt resistor. This hotend would heat up to 225 before stopping altogether when set to 185 for PLA.
I had the problem with plastic oozing from the threads of the acorn nut, but I fixed it with high temp gasket silicone.

The resistor now completely failed and is showing 1M ohm resistance.

Does using a 1 Ohm resistor screw up anything in sprinter causing the resistor to heat up too much? Is there anything i can change in Sprinter settings?
I am thinking of trying 1" brass this time with the same length ptfe.


Regards,

I'm using 2x 6.8ohm wire-wound resistors in parallel. The PID settings of the firmware should adjust automatically for different heater wattages.

If you use 1Ohm and 12V you get 144W in a 5W resistor, so no wonder it overshoots and then burns out, That is about the power we use for a heated bed. An extruder only needs about 20W.

---

[www.hydraraptor.blogspot.com]

Thanks for the clarification nophead and aplavins.
I have ordered 5ohm 10watt aluminum encased resistors just in case of this situation.
PWM does not seem to compensate for the actual current flowing thru the resistor causing the overheat issue. Oh well.

if your tight on money like i have been in the past, you can use a bit a broom handle with a hole drilled in it instead of the ptfe block, it will only last maybe day of two but it's a good way to get out of trouble

re, Mendel90_hot_end.png

I think if the heating core was tapped for M6,
then it could be located nearer the nozzle, and
only one M6 locking nut would be needed against
the top; with maybe a serrated washer to lock
against effects of highs/lows in temperatures.

Also, a more balanced core around the nozzle
would make for more even heat distribution with
a second thermistor.

Edited 1 time(s). Last edit at 08/13/2012 04:20PM by Scrachbuilder.

Yes I have made newer ones with a single locknut and a spring washer above the block.

I don't understand why it needs a second thermistor?

---

[www.hydraraptor.blogspot.com]

I noticed that when I use 5Ohm 10 watt resistor, hot end temperature varies between 185 and 205C. With 12 volts, I understand that load on 5Ohm goes to 28Watts. If I add 4Ohms of resistors(to keep it below 10Watts on 5 Ohm) in series away from the hotend (due to size restriction), it takes forever to heat up the hot end (5C per min). What is the right balance for sticking to 5C range around 190C with a decent heating gradient?

I am generously using kapton tape around the copper block and aluminum clad 5Ohm resistor, so convection cooling is negligible.
I am going to try dropping the external resistors one at a time and retest. I just wanted to know if there are any other variables I am not looking at in Sprinter configuration.

Edited 1 time(s). Last edit at 08/17/2012 06:59PM by peddiparth.

An aluminum clad resistor is overkill for the hot end really. As long as the resistor is decently thermally coupled to the hot end, the controller keeps the resistor from getting too hot and burning out. A single 5Ohm 3W wire-wound resistor works just fine for the hot end. The wattage rating of the resistor is to keep it from getting too hot, but our extruder control does this for us.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Cameron, what is your temperature range when you set it to (say) 185C?
Does your hot end stay at 185 exactly, or does it overheat and overcool giving you a temperature range?

peddiparth Wrote:
-------------------------------------------------------
> Cameron, what is your temperature range when you
> set it to (say) 185C?
> Does your hot end stay at 185 exactly, or does it
> overheat and overcool giving you a temperature
> range?

Using PID control, the overshoot when heating to 200C is only about 2 degrees. After which it maintains 200C +0-1 degrees. So the range is only 199 to 200C during printing, and a max of 202C when first warming up, when I print at 200C.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

How fast does yout hotend reach 200C?

Cameron, please share your PID settings of Sprinter so that I can fix my settings.

peddiparth Wrote:
-------------------------------------------------------
> Cameron, please share your PID settings of
> Sprinter so that I can fix my settings.


The PID settings depend heavily on the hotend. I use the PID autotune functionality of Marlin (M303 SXXX). You could even upload Marlin, do the PID autotune, write down the values, then re-upload your Sprinter firmware if you wanted.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Switched to Marlin firmware. Printing at 190C (+-3C) is no longer a problem

I wrote this months ago. I acquired a bunch of material for experimentation but haven't been able to work on it. So, I thought I should stop sitting on the idea and contribute it to the RepRap community. Might be a really bad idea, or might have merit. Either way, I hope this will make for some interesting discussion.

Motivation
++++++++
I’ve been wondering about the possibility of using infrared semiconductor lasers (or LEDs) as the heat source in the hot end of an extruder. The speed at which the heat can be applied and removed seemed to be appealing particularly with respect to nozzle oozing.

I searched the reprap forums and elsewhere to see if someone had tried to use lasers for ABS (or PLA) filament melting but wasn’t able to find anything definitive, or informative. Hopefully, this post will provide some meaningful data for those interested in the possibility of using lasers in the hot-end of an extruder.

Assumptions
++++++++++
I make the assumption that the filament is contained in a glass tube and the IR laser must heat the ABS filament through the glass.

I’m NOT assuming that using IR lasers as the heat source in an extruder is practical in any way. This is really just a theoretical and experimental exercise to satisfy my curiosity. I’m specifically ignoring application issues such as user safety, cost, construction, component sourcing, weight, etc.

Please feel free to correct any errors I have made.

Energy Formula
++++++++++++
Energy Required = DeltaT * Mass * Specific Heat + Mass * Latent Heat

ABS material properties
++++++++++++++++++
I am deliberately trying to use conservative values so that the calculated energy required to melt ABS will be worst case.

ABS is Acrylonitrile Butadiene Styrene

Specific Heat = 2.1 KJ/Kg-DegreesK = 2.1 J/g-DegreesK
Latent Heat = 0 (ABS is an amorphous material)
Density = 1.25 g/cm3

ABS melt temperature = 105 DegreesC
ABS becomes soft at about 90 DegreesC

Other Parameters
++++++++++++++
Maximum feedrate = 90 mm/s (high)
Extruder diameter = 0.5 mm

NOTE: at a given feed-rate (i.e. 90 mm/s) a larger diameter nozzle extrudes more mass than smaller diameter nozzle in a given time (i.e. 1 second), and therefore requires more energy.

Calculations
++++++++++
Area of Extrusion = PI * D^2 / 4
Area of Extrusion = approximately 0.2 mm^2

Volume of Extrusion / second = Feedrate * Area of Extrusion
Volume of Extrusion / second = 90 mm/s * 0.2 mm^2 * ( 1 cm / 10 mm)^3 = 0.018 cm^3 / s

Mass of Extrusion / second = Volume of Extrusion / second * Density
Mass of Extrusion / second = 0.018 cm^3/s * 1.25 g/cm^3 = 0.0225 g/s

Specific heat * Mass of Extrusion / second = 2.1 J/g-DegreesK * 0.0225 g/s
Specific heat * Mass of Extrusion / second = approximately 0.05 J/s/DegreesK

1 J/s = 1 Watt
A temperature Delta of 1 DegressK = Delta of 1 DegressC

0.05 W/DegreesK… or … 20 DegreesK/W
0.05 W/DegreesC… or … 20 DegreesC/W

90% IR absorption efficiency by ABS
90% IR glass transmission efficiency

90% * 90% = approximately 80% of IR optical power from laser, through glass, into ABS filament

20 DegreesC/W * 80% = 16 DegreesC/W (when filament is behind glass)

@ 90 mm/s feed-rate
++++++++++++++++
If filament is pre-heated before laser heating to melt filament

89 DegreesC pre-heat + 1 W laser @ 16 DegreesC / Watt = 105 DegreesC melt temperature
73 DegreesC pre-heat + 2 W laser @ 16 DegreesC / Watt = 105 DegreesC melt temperature
57 DegreesC pre-heat + 3 W laser @ 16 DegreesC / Watt = 105 DegreesC melt temperature
41 DegreesC pre-heat + 4 W laser @ 16 DegreesC / Watt = 105 DegreesC melt temperature

If filament is at Room Temp (no pre-heating)
25 DegreesC + 5 W laser @ 16 DegreesC / Watt = 105 DegreesC melt temperature

NOTE: Less optical power is needed as slower feed-rates.

Conclusions
++++++++++
The calculated laser power required is dangerous particularly given that an IR laser is invisible.

However, the calculated laser power isn’t so high as to rule out the possibility of using IR lasers to melt ABS filament altogether. It should be practical to do some bench experiments in the lab with the right precautions and safety equipment.

Pre-heating the filament will reduce the laser power required (obvious, but worth noting).

Power decreases with reduced feedrate.

Power decreases with reduced extruder diameter.

We don't just need to get the filament to melt, we need to get it hot enough to bond with the previous layer. Temperatures over 250C are common. That brings your needed power up to around 14W to heat it from room temperature.

For some time I've been trying to find information about the viscosity of melted ABS. I kept searching on viscosity but tonight I discovered "material flow rate". My first couple searches turned up the following:

ASTM D21398 Standard Test Method for Melt Flow Rates of Thermoplastics by Extrusion Plastometer

[www.ineos.com]

And this [plastics.ides.com]

I also found this plain english explanation of "Melt Flow Index" from here...

Cabot Melt Flow Index

One of the properties most often quoted for both natural polymers and masterbatches is Melt Flow Index (MFI). What is MFI and why is it significant ? Melt Flow Index is the output rate (flow) in grammes that occurs in 10 minutes through a standard die of 2.0955 ± 0.0051 mm diameter
and 8.000 ± 0.025mm in length when a fixed pressure is applied to the melt via a piston and a load of total mass of 2.16 kg at a temperature of 190°C (some polymers are measured at a higher temperature, some use different weights and some even different orifice sizes).

Melt Flow Index is an assessment of average molecular mass and is an inverse measure of the melt viscosity; in other words, the higher a MFI, the more polymer flows under test conditions. Knowing the MFI of a polymer is vital to anticipating and controlling its processing. Generally, higher MFI polymers are used in injection moulding, and lower MFI polymers are used with blow moulding or extrusion processes.

And then I found Wikipedia's Melt Flow Index

I've seen numerous posts in the RepRap forums lamenting the absence of data on ABS material properites so I thought I should post these tidbits.I didn't know where to post this info so I dropped it here. This info might already exist elsewhere in the forums but I didn't encounter it.

TC