Tape Lights and 5mm LEDs are Not the Right Approach to Light Conditioning of SLA 3D Printed Parts

Posted: February 1, 2017 in Light Cure
Tags: , ,

SLA printing uses laser light to cure resin in a bath to generate 3D parts, one layer at a time. The finished parts are smooth, with finer detail than what can be accomplished using (Fused Deposition Modeling), which extrudes plastic that is deposited in layers to build a part on a build platform. However, SLA printed parts require post processing to make them usable, even for model creation. Each part required excess uncured resin to be cleaned away, usually in an alcohol bath. Further, to eliminate surface stickiness and improve overall strength, it is necessary to use light cure conditioning to complete the curing of the finished model.

For desktop printer owners, the search for an economical post-process light cure solution leads to low-cost solutions. One such solution is to purchase tape light strips. These are inexpensive and operate from simple low voltage power supplies. They can be purchased from e-bay for a few dollars, and proclaim high-efficiency and violet light. Most use a standard mid/low power 5050 LED package and deliver anywhere from 395 to 410nm of violet light.  Another approach, which we have attempted, is use of 5mm LEDs in arrays on a simple custom circuit board.

Before I go any further, I will state now that after extensive testing and experimentation attempting to discover a super-low cost light cure conditioning solution, none of the low-end, low power light source solutions will work effectively. They will require very long cure times, resulting in minimal improvement in part strength. While they may, after hours of cure time, have improvement in reducing surface tackiness, for the most part, these approaches are ineffective in penetrating beyond the very surface of the material. In the case of black, toughened, or pigmented parts, the effectiveness is even less.

This leads to an important number. 30. To achieve a reasonable cure, at a depth required to impart any improvement in part strength, demands a minimum of 30mW/cm². That is 30 micro-watts per CM² on the surface of the part. This was derived from numerous empirical tests, using a Barber Coleman Impressor to measure hardness of samples ranging from .030″ to .25″, on materials ranging from clear to black, toughened and cast-able. This included flexible materials as well, using tear and tear testing, rather than hardness testing, which is irrelevant for that material. I found that anything less than 30mW/cm² delivered less improvement, while power levels greater than that began to follow a diminishing return. However, for very thick parts, >.188″, having as much as 50mW/cm² can be advantageous.

The power level of 30mW/cm² is part of the formula. Another is time. The smaller the power level, the greater the exposure time required, assuming the power level present is enough to create a reaction in the first place. This leads to another metric, joules. Joules represents the actual energy required to complete a cure. It is a factor of power (Watts) over time (Hours) x 3600. For use of mW and minutes, the formula is Minutes x mW x 3600. 1mW x 1 minute = .06 joules. Assuming we achieve a full cure in 15 minutes, using a source delivering 30mW of power to the target surface, the energy involved is a total of 27 joules. This energy must be delivered to the resin located deepest within the part itself (core) and furthest from the light source.

Thicker and darker parts require more, and thin, transparent parts less energy. The issue is transmission losses. Thick pigmented parts block light from penetrating deeper. From our testing, the 30mW minimum will parts made from pigmented (black or dark gray) materials fully. If the transmission loss is 50%, from the outer skin of the part to the center, the center will take twice as long to cure as the skin. This is included in the 30mW/cm² stated, as the final state of “cure” was when the sample parts were cured through. We also tested pure resin in its liquid state, measuring how deep the cure reached over time. The minimum amount of power required to keep the cure reaction alive differs from resin to resin. However, it appears that when power levels drop to below 1.2mW/cm², the cure reaction stops and regardless of the exposure time – meaning there is no additional cure benefit.

Without getting into too much more detail, let’s use 30mW as a baseline to achieve a 15 minute cure (27 joules) for most parts, and move forward.

How does it related to the tape light and through hole 5mm LED sources? It all comes down to efficiency. Not lumen/Watt efficacy, but plug efficiency. This is a measure of how may radiant energy Watts are emitted for each Watt input into an LED. A typical 5mm LED, and most low power SMD products operating at 3.5V deliveirng 405nm, operating at 20mA, generates roughly 7.5 mW of radiant energy. That’s .07W in, and .0075W out, for an efficiency of 10.7%. This includes the 5050 package LED (which applies to other sizes, including 3020, 3528 etc.) common to tape light products. These low power LEDs operate on tape systems with no thermal path or component. They are also used because they are cheap, usually costing a few cents each in bulk. These components are simply not powerful enough to be used for any light-cure processes, including SLA light conditioning. I will explain why here.

Moving upward to mid power products (+/- 1W), typically in a round body SMD LED, delivers ~440mW at 350mA x 3.5Vf, for 27% plug efficiency – and cost around $3.00 each. A higher quality 3535 package SMD component can be found that operates at 500mA, at 3.4Vf, delivering 700mW or radiant energy for 41% plug efficiency. These devices cost around $3.80 each in bulk,  High performance high power packages can deliver as much as 1500mW, at 700mA x 3.5V, for an efficiency of 61%. These devices will cost around $5.00 each in bulk. We use the high performance packages in our commercial cure light systems for this reason. All of these devices must be attached to some form of thermally conductive surface to dissipate their heat energy, which means circuit boards with thermal vias mounted to heat sinks (or metal panels), or MCPCBs. Most of these devices are suitable for light curing and SLA part post-print conditioning.

Getting back to the UV light conditioning systems… To get an idea of the power needed to facilitate a true cure state or conditioning effect, let us assume that all of the LEDs in a comparison have a distribution of 120 degrees, and that 50% of that falls into a usable zone, and that the distance from the light source to the target is uniform, which demands an additional increase of source power of 4X to overcome the inverse square law loss. That means that all of the sources are subject to an optical efficiency multiplier in application of .50 x .25, or .125.  If this is applied to a 1CM area, the power required to attain a target 30mW/cm² is 240mW at the source. To cover a target area of 25cm² from four sides, would demand a total of 24,000mW of total power. For comparison sake, this is the central area of a box measuring 10″ x 10″, lighted from four sides.

If I apply the various LEDs to this, here is what I get:

  • Using tape light LED products delivering 7.5mW per LED requires a total of 3,200 LEDs (1,600 feet id the LEDs are space .5″ apart). The total wattage of this system would be 224W
  • Using mid power LEDs delivering 440mW per LED requires a total of 54 LEDs. The total wattage of the system would be 132W
  • Using standard performance high power LEDs delivering 700mW would require 34LEDs and consume 59.5W
  • Using the highest performance high power LEDs available, delivering 1500mW would require 16 LEDs and consume 39W

If we accept the lower power level of the LEDs and add time to compensate for their deficiency, here is what we get. These assume a presence of 24,000mW of total power  required to attain the average 30mW on the target surfaces, delivering a 15 minute cure cycle (27 joule ideal goal).

  • Using a more reasonable count of 160 LEDs from a tape light or 5mm LED source, we will realize 1,200mW total (average of 1.5mW/cm²). Using the previous 24,000mW power  to attain the average 30mW on the target surfaces for a 15 minute cure cycle (27 joule ideal goal), the additional time required will be 300 minutes, or ~5 hours. The real issue here is part thickness. If the part is made from dark resins, or is thick (>.125″ section) the transmission loss is going to be so great that the minimum power required to cause a cure reaction will not be attained, regardless of the time of exposure.

For perspective of costs (light source alone, no cabinet or box) the implications are as follows:

  • 250 5mm LEDs at $.70 each would cost $175 plus the cost of circuit boards and power supplies (+/- $300) and require a 5 hour cure cycle – unlikely to cure parts >.125″ thick
  • 62 feet of tape light (250 LEDs) would cost  $434 ($7 per foot) plus power supplies to support 224W load (+/- $500) and require a 5 hour cure cycle – unlikely to cure parts >.125″ thick
  • 54 mid power LEDs at $3.00 each would cost $162.00 plus circuit boards and driver ($+/- $500 total cost) and deliver a 15 minute cure
  • 34 standard high power LEDs at $3.80 each would cost $129.00 plus cost of circuit boards and driver (+/- $420 total cost) and deliver a 15 minute cure
  • 16 high performance high output LEDs at $5.00 each would cost $80, plus boards and driver (+/- $320 total cost) and deliver a 15 minute cure

Obviously, the “cheap” approach of using tape light to create the cure box is not actually that cheap, and is not an effective solution. While some will argue they have built systems using tape inside wood boxes effectively, I will argue that the proclaimed results are anecdotal at best. Yes, the parts may be a little less tacky after five hours exposure, but they are not significantly stronger. I know this from direct experience.

Our own first effort at a light box employed 125 5mm LEDs operating at their max of 30mA. The measured interior radiant energy falling onto the part inside was 1.63mW/cm². Even with cure times as long as 5 hours, the only usable result was on thin parts, and a general reduction in surface tackiness for heavier models. For this reason, we stopped our effort to package this particular product for sale. That does not mean we quit. Far from it. Our next step was to build a truly effective cure box for SLA parts and small assemblies adhered with UV cure adhesives.

The next installment on this topic will be our solution. It will deliver >30mW/cm² on the target surface, provide enough room for parts and assemblies as large as a 10″ cube, and offer an affordable solution over cobbled together tape light solutions that will cost as much or more, delivering less effective end results. Stay tuned!




Leave a Reply