Seven-tank pretreatment chemistry: what each bath does

A titration that reads inside its window proves one thing: the bath contains the chemistry the sheet says it should. It does not prove the bath will phosphate a part. Two lines can titrate to the same free acid and produce coatings that differ by a factor of two, because one has 2 g/L of dissolved iron and a floating oil layer and the other does not.

Seven-tank pretreatment chemistry is therefore managed as two separate problems. The first is concentration, which titration answers. The second is contamination, which titration cannot see, and which decides the coating weight, the adhesion and the salt-spray hours just as surely. This article takes the baths one at a time, works out a real dose, and sets the point at which a bath is dumped rather than corrected.

What does each bath in the seven-tank line actually control?

Each bath controls one reaction with a concentration, temperature and dwell window. Move outside any one and the coating changes, even when the tank is present, the pump is running and the part looks clean.

Bath 1 is an alkaline cleaner. Its job is to saponify drawing oils and lift particulate soil off the steel, and it does that through free alkalinity, surfactants and temperature. Bath 2 is a rinse whose only job is dilution: it exists to stop cleaner being carried into the conditioner. Bath 3 is a conditioning stage that seeds the surface with titanium colloid so the conversion crystals nucleate densely and finely rather than in a coarse, patchy layer.

Bath 4 is the conversion stage. In a zinc phosphate bath, free acid dissolves a small amount of iron from the surface, the dissolved iron raises the local pH at the interface, and zinc phosphate crystallises out onto the steel. Total acid is the reservoir that keeps that reaction supplied; free acid is the attack that starts it. Bath 5 rinses the soluble phosphate and acid away. Bath 6 is a demineralised rinse that removes the last of the dissolved salts before drying. Bath 7 is dry-off, where the part leaves the line dry at the jig contact and inside every return.

The chemistry that follows the substrate is not negotiable. A zinc phosphate bath that runs CRCA well will coat galvanised steel poorly unless the cleaner and the acid attack change, and it will not run aluminium: dissolved aluminium ties up the fluoride that controls the attack rate.

The titration numbers and how a dose is calculated

Every addition starts with a difference between the measured titre and the set point, converted into a mass of chemical for the tank volume. The arithmetic is simple and it is the part most often done by eye. A dose worked out from a titration is reproducible; a dose worked out from a jug is not.

Free alkalinity in the cleaner. Take a 10 mL sample, titrate with 0.1 N hydrochloric acid to the bromocresol green endpoint, and express the result as a percentage of the product. A 3,000 L tank reading 1.6% against a 2.0% set point is short by 0.4 percentage points, which is 0.4% of 3,000 kg of solution, or 12 kg of cleaner. At a product density of about 1.1 kg/L that is roughly 11 L added into circulation, not poured onto the surface.

Free acid and total acid in the phosphate bath. Take 10 mL, titrate free acid to bromocresol purple and total acid to phenolphthalein, and read both as points. A bath at 0.8 points free acid against a 1.0 point set point on 3,000 L is short by 0.2 points, which is 0.002 equivalents per litre, or 6 equivalents in total. Neutralised with 32% sodium hydroxide at about 10.6 mol/L, that is 0.57 L — a small addition, which is why the addition is made diluted and under agitation rather than dumped into one corner of the tank.

Accelerator. Nitrite or another accelerator is measured separately, usually by a simple spot or titration method, because it is consumed by the reaction itself rather than by drag-out alone. A bath can hold perfect acid numbers and still coat slowly if the accelerator has fallen away, and the symptom is a coating that is thin and grey rather than a bath that looks wrong.

These three calculations scale with tank volume and are checkable. An operator who adds "about a bucket" to a 3,000 L tank adds a variable the next shift inherits.

The seven baths, their set points and their checks

The table gives working ranges for an alkaline-cleaner, zinc-phosphate line on CRCA. Read them as a starting point that the approved plant process and the customer specification narrow for a specific part; aluminium, galvanised steel and heavy-gauge fabrications all shift the numbers.

Two rows in that table carry more weight than the rest. Bath 4 decides the coating weight, and bath 6 decides whether the coating survives. Everything between them exists to protect those two.

Contamination is the variable the titration cannot see

Iron is the first contaminant that matters. The conversion reaction dissolves steel, and the dissolved iron stays in the bath until it is either precipitated as sludge or dumped with the bath. The loading can be estimated before it becomes a problem. A line that treats 8,000 m² of steel a month at roughly 1 g/m² of iron dissolution puts about 8 kg of iron into the bath per month. In a 3,000 L tank that is 2.7 g/L per month, so a bath reaches 4 g/L in about six weeks and 5 g/L in seven.

Past that point coating weight falls and crystals coarsen, because the free acid works against a bath already saturated with the metal it dissolves. Sludge volume climbs too, and sludge not drawn off the tank bottom is picked up by the pump and redeposited as a grey speckle.

Oil is the second contaminant. Alkaline cleaning emulsifies drawing oil, and the emulsion has to be broken and removed at the surface weir. A tank that is not overflowing has no route out for the oil: it accumulates, the cleaner saturates, and the bath starts leaving an oil film on the part at the moment it leaves the tank. The titration still reads in range because the alkali is still there; it is simply attached to the wrong thing.

Aluminium is the third, and it is specific to mixed-metal shops. Aluminium dissolves in an acid phosphate bath and ties up fluoride, which is the ion that keeps the attack rate stable. Above roughly 50 mg/L of aluminium the coating becomes patchy and slow, and the correction is not an addition — it is a dump. This is the reason aluminium parts normally run a chrome-free zirconium or titanium conversion instead.

Temperature control is a heat balance, not a set point

A controller holding a tank at 55 °C is holding the temperature of the thermocouple, which sits near the heat exchanger inlet. The part sees the temperature of the volume it enters, and every cold hanger takes heat out of that volume.

Take a 3,000 L bath and a 200 kg steel hanger at 25 °C entering at 55 °C. Heating the hanger through 30 °C takes about 2.9 MJ at a specific heat of 0.49 kJ/kg·K. Removing that energy from 3,000 kg of water at 4.19 kJ/kg·K drops the bath by about 0.23 °C. Sixty hangers in a shift remove 174 MJ, which averages about 6 kW of reheat across eight hours before losses and evaporation.

That is why a line running continuous loads shows a sawtooth on the chart and a batch line shows a flat line with a dip at the start of every batch. Neither is wrong if the bath recovers before the next part enters. A bath that never recovers runs the shift at the bottom of its window and coats thin.

When to correct a bath and when to dump it

The decision rule that keeps a line stable is short enough to put on the process sheet.

• Correct when the titre is within about 15% of set point, iron is below 3 g/L, the oil layer is being removed at the weir, and the last coating-weight check passed.
• Correct and investigate when the titre is in range but coating weight has fallen: check accelerator, iron, conditioner age and rinse conductivity before touching the acid balance.
• Dump when iron passes 4–5 g/L, when aluminium passes roughly 50 mg/L, when three additions in one shift fail to hold the titre, or when coating weight stays below specification after a correct addition and a confirmed accelerator reading.
• Dump immediately when the bath has been contaminated with the wrong chemistry, when sludge is being drawn into the pump, or when the tank has been idle long enough for biological growth or stratification.

A planned dump is cheaper than a month of parts that fail salt spray. The spent bath goes to the effluent treatment plant, not the drain.

What the record has to show

A pretreatment record has to show that the parts made in a shift were made under the conditions the process sheet requires. That means a time, limits printed next to the reading, a signature, and coverage of every shift that ran production.

In practice that means a titration sheet per bath with the set point, the minimum and the maximum printed on it, not a blank column. It means a conductivity reading on bath 6 every shift, because that is the number that shows a DI plant falling behind. It means iron and aluminium on a weekly or monthly frequency matched to the bath's loading, and a coating-weight coupon tied to a batch number whenever the drawing requires one. And it means a dump log with the date, the reason, the volume and the ETP batch that received it.

The failure this prevents is drift: a bath corrected by feel for three weeks, a rinse that stopped overflowing when a valve was left closed, a conditioner unchanged since the last shutdown. Each produces parts that look right and fail later, and each is visible in a record that carries limits.

How Ganesh runs pretreatment chemistry

Ganesh runs a seven-tank pretreatment line ahead of an 18 ft × 14 ft electrostatic powder booth, with a 16 ft × 16 ft shot blast chamber for scale and weld discolouration on heavier fabrications. Parts from the press shop, the laser and the three CLOOS robotic weld cells converge on the same line, which is what keeps a coating consistent when the assembly contains a pressed bracket, a laser-cut panel and a welded frame.

Baths are titrated on the shift, rinse conductivity is read on the final rinse, and dry film thickness is checked with a DFT gauge with adhesion confirmed by cross-hatch test. First article inspection and PPAP documentation are available where a programme requires them, and the records sit inside the ISO 9001:2015 system. The line is served by an effluent treatment plant that operates under ISO 14001:2015, at the Mumbai (Vasai) plant. The paint shop capability page covers the route from blast to oven.

The plant, quality and sustainability hub indexes the articles on calibration, inspection and environmental control that surround this chemistry. If the chemistry is stable and the coating still fails outdoors, the next thing to read is why powder coating fails on outdoor enclosures, which separates a pretreatment problem from a cure or film-build problem.

What the drawing and the RFQ have to carry

The line can only run what the part and the specification allow. Six fields on the drawing or in the RFQ pack remove most of the chemistry guesswork:

1. Substrate and grade, because CRCA, galvanised steel and aluminium do not share a bath chemistry.
2. The service environment and the corrosion class the finish has to meet, so the coating weight target is set before quotation rather than after a failure.
3. Any standard the finish must be tested to, with the test method named: coating weight to ISO 3892, salt spray to ISO 9227, adhesion by cross-hatch.
4. Drainage: the low-point hole in every closed section, and any preferred hanging orientation.
5. Masked features: threads, earth points, mating faces and bores that must stay conductive.
6. Whether the part is a mixed-metal assembly, because a steel bracket riveted to an aluminium panel cannot be run through a phosphate bath without contaminating it.

A drawing that names the substrate and the test method is quoted faster and coated more consistently than one that says "powder coat, colour as per sample". The seven-tank pretreatment line walkthrough covers the physical path through the tanks, and phosphating and conversion coatings covers what the conversion layer does under the paint.

When this chemistry is the wrong route

Seven-tank zinc phosphating is not a universal answer, and the limits are worth stating at quotation. A fabrication larger than the tank opening cannot be dipped, so it moves to a spray pretreatment with a different sequence and a different set of controls. A closed assembly with trapped volumes will carry bath liquid through every stage and release it during cure. Aluminium runs a chrome-free conversion rather than zinc phosphate. Galvanised steel needs a cleaner and an acid attack that do not strip the zinc, which is a different bath, not a different tank count.

Where the part cannot be dipped, the answer is a different route or a design change that lets it drain. A sealed box section run through the line carries liquid inside it and blisters from the inside out.

Send the drawing with the substrate, the service environment and the test method the finish must meet, and the engineering team will return the bath sequence, the coating weight target and the record the line will sign. The quote request accepts PDF, DXF, DWG and STEP files.