Solder and Molten Metal Measurements
Solder and hot dip metal coating measurements can be carried out in the FTA heated Environmental Chamber. This note discusses some of the characteristics and features of such measurements.
Maximum Temperature of Chamber
We refer to the measured wall temperature of the chamber in the following. This is what is displayed on the controller. The maximum temperature the chamber can reach is determined by several factors, listed in importance from most important to least:
AC line voltage. The line voltage is a major influence. In a 23C lab, the following maximum temperatures were observed:
- 100V: 242C
- 105V: 258C
- 110V: 271C
- 115V: 290C
- 120V: 309C
These temperatures were measured after 1 hour of heating. The temperatue rises quickly when the heaters are first turned on, then tapers off to an equilibrium value at the end. The equilibrium temperature is determined by the heater output (set by the AC line voltage) and the losses from conduction and radiation (discussed below).
Bottom plate tight. It is assumed the two thumbscrews holding the bottom plate in place are tight, otherwise the bottom will not be in good thermal contact with the rest of the chamber.
Conduction losses to main frame. A thin 1/4" (6mm) ID stainless steel washer should be inserted between the chamber's stainless steel mounting plate and the main vertical plate of the instrument. This reduces heat loss to the frame.
Conduction losses to the air. The Kevlar fabric cover should be used to insulate the chamber, both to minimize heat loss and to protect the user against accidentally touching the hot stainless steel cover.
Conduction losses to the two atmosphere ports (barbed fittings for 3/16 or 4mm ID tubing). These ports can be plugged or removed and the holes filled with #10-32 screws to minimize air transfer. Clearly any purge gas or other gas flow through the chamber will remove heat and lower the maximum temperature. The above maximum temperature measurements were made without this precaution, however.
Conduction losses to the needle port. The needle port should be plugged if a needle is not used. The above maximum temperature measurements were made without this precaution, however.
Note the chamber is rated for 120VAC service. A voltage-adjust transformer should be employed if the local line voltage is below 120V (delivered to the input of the controller -- the heater itself will be a volt or so lower, but this is accounted for). No particular harm will come if a voltage adjust transformer is employed and the applied voltage is a volt or two above 120V.
The following temperature rise was measured at 120V line voltage:
- t=0 min: 23C
- t=15min: 208C
- t=20min: 246C
- t=30min: 272C
- t=45min: 298C
- t=60min: 309C
Lower line voltages will cause temperature rise at a proportionally lower rate.
Cool Down Rate
The cool down time is quite long, similar to the heating time, so the use of a small electric fan to blow air on the chamber is recommended.
Actual Sample Temperature
During heat-up periods, the sample lags behind the chamber temperature because thermal transfer to the sample, sitting loosely on the stage, is poor. This is a particular issue when you want to melt the sample because a great deal of heat must be transferred to the sample to cause the phase change. This heat transfer is inefficient (slow) because of the light contact between the sample and stage. Most of the heat must be conducted through the atmosphere within the chamber and by radiatiuon from the chamber walls. Except in the case of melting a sample, the sample is about 10 minutes behind the wall temperature. Whatever the wall temperature is now, that's what the sample will be in 10 minutes.
The sample also is at a slightly lower temperature, particularly at high sample temperatures, because of radiation from the sample. The sample sits in a region where it can "see" three windows: the observation window above it, the illumination window behind it, and the microscope viewing window in front of it. Experiments made by blocking these windows and measuring the temperature rise have shown that there is noticeable loss through the windows. Unfortunately we need the windows to do our work. Experiments have been made on the freezing point of molten solder and comparing that with the observed wall temperature. Standard electronic solder (63% SN, 37% Pb) has a published melting point of 182C. This solder was melted by raising the chamber to a high temperature and then letting it slowly cool down. The solder was probed by a wire inserted through the needle port. The point at which the molten solder solidified occured when the wall temperature was about 197C, implying the sample was roughly 15C below the wall temperature. The bottom plate was separately measured and found close to the wall temperature, so the only logical reason for the sample being lower is radiation loss. Note it was possible to melt solder against the chamber walls and bottom, even when the sample solified. This shows there was truly a difference.
How to Melt Solders and Similar Metals
On a practical basis, you can not melt solders unless you use a flux. Without a flux, you have the following chain of events:
The chamber heats slowly (say an hour).
The outside surface of the solder or metal oxidizes. You can see the tarnish (change of color) on the surface.
Because the solder or metal is typically cut into small pieces, the volume and mass are low. Even if the inside melts, the outside oxidized surface is likely to maintain its shape. We have seen cases where the sample deforms but never "melts" in the expected fashion, even when the actual sample temperature is 50C above the nominal melting point.
To melt solders without the use of a flux, the temperature must be way above the melting point so the agitation of the molten metal inside breaks down the oxidized skin.
If you place an electronic solder wire with internal flux in the chamber, the internal flux will not escape the wire and effectively you have no flux. When you use electronic solder wire in its normal fashion, you break the flux loose my mechanically rubbing the solder end to the hot work. You can not do that inside the chamber.
When we use a flux, the situation is entirely different. We successfully ran a set of pure tin (Sn, melting point 232C) wettability tests on copper plates. We ran approximately 50 tests in two days, using the following protocal:
The sample and the tin pellet were dipped in flux before placement on the chamber stage. The stage would accommodate three samples at the same time, so three could be run with one thermal cycle.
Line voltage was 120V.
The controller was set to 300C so the chamber would heat as rapidly as possible.
The time at which the chamber wall temperature reached 232C was noted. This was typically somewhere between 15 to 20 minutes after starting.
About ten minutes after reaching 232C, the chamber wall temperature would reach 270C and the sample pellet would melt and spread on the copper plate to make a contact angle the way you would expect. Sometimes this was a little sooner (lower temperature) and sometimes a little later (higher temperature), depending on the treatments made to the plate. This was the object of the study. From the above discussions, we note:
- The sample will lag behind the wall temperature as the wall temperature rises.
- The sample must absorb extra heat to actually melt, and this takes time.
- The sample will lose heat to radiation during the whole process.
After the contact angle images were captured, the controller was turned off and the cooling fan turned on. Approximately 15 minutes later the chamber was cool enough to unload the old samples and load new ones. The total turn-around was a little better than 3 samples per hour or 24 per 8 hour day.